Overview
Comment: | wip |
---|---|
Downloads: | Tarball | ZIP archive | SQL archive |
Timelines: | family | ancestors | descendants | both | v1.70-refactor02 | v1.70-defunct-try |
Files: | files | file ages | folders |
SHA1: |
5d0accfc514ca948958ca98cf6586e83 |
User & Date: | matt on 2019-12-30 22:57:35 |
Other Links: | branch diff | manifest | tags |
Context
2019-12-30
| ||
22:58 | wip check-in: 8a7e2c859e user: matt tags: v1.70-refactor02, v1.70-defunct-try | |
22:57 | wip check-in: 5d0accfc51 user: matt tags: v1.70-refactor02, v1.70-defunct-try | |
22:34 | wip check-in: 05c74be94c user: matt tags: v1.70-refactor02, v1.70-defunct-try | |
Changes
Modified Makefile from [844cc54c6a] to [39870f25b1].
︙ | |||
184 185 186 187 188 189 190 | 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 | - + + - + | mofiles/tasksmod.o mofiles/odsmod.o mofiles/tasksmod.o : mofiles/commonmod.o mofiles/pgdbmod.o mofiles/rmtmod.o : mofiles/commonmod.o \ mofiles/apimod.o mofiles/ulex.o mofiles/itemsmod.o mofiles/apimod.o : mofiles/dbmod.o mofiles/commonmod.o mofiles/servermod.o mofiles/runsmod.o : mofiles/testsmod.o mofiles/mtmod.o mofiles/mtmod.o : mofiles/mtconfigf.o mofiles/rmtmod.o mofiles/tasksmod.o \ |
︙ |
Modified docs/code/module-hierarchy.dot from [15eb538f75] to [fbbdb982aa].
︙ | |||
65 66 67 68 69 70 71 | 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 | - + - - - - - - + + + + + + + + | // mtmod mtconfigf -> mtmod; commonmod -> mtmod; rmtmod -> mtmod; servermod -> mtmod; taskmod -> mtmod; mtargs -> mtmod; |
Modified launchmod.scm from [07359d2b62] to [fa7416b0c1].
︙ | |||
20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 | 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 | + + + + - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - | (declare (unit launchmod)) (declare (uses commonmod)) (declare (uses rmtmod)) (declare (uses mtconfigf)) (declare (uses testsmod)) (declare (uses mtargs)) (declare (uses subrunmod)) (declare (uses dbmod)) (module launchmod * (import scheme chicken data-structures extras files) (import (prefix sqlite3 sqlite3:) posix typed-records srfi-18 srfi-69 format ports srfi-1 matchable z3 (prefix base64 base64:) regex call-with-environment-variables csv) (import commonmod) (import rmtmod) (import (prefix mtconfigf configf:)) (import testsmod) (import (prefix mtargs args:)) (import subrunmod) (import dbmod) ;; (use (prefix ulex ulex:)) (include "common_records.scm") (include "db_records.scm") (include "run_records.scm") |
Modified mtmod.scm from [2807629944] to [6220786280].
︙ | |||
25 26 27 28 29 30 31 32 33 34 35 | 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 | + + - + + + + + + + - + + + + | (declare (uses testsmod)) (declare (uses tasksmod)) (declare (uses mtargs)) (declare (uses dbmod)) (declare (uses pgdbmod)) (declare (uses servermod)) (declare (uses launchmod)) (declare (uses subrunmod)) (declare (uses stml2)) (module mtmod * |
︙ | |||
1087 1088 1089 1090 1091 1092 1093 | 1098 1099 1100 1101 1102 1103 1104 1105 1106 1107 1108 1109 1110 1111 1112 | - + | (let ((db (cdr *task-db*))) (if (sqlite3:database? db) (begin (sqlite3:interrupt! db) (sqlite3:finalize! db #t) ;; (vector-set! *task-db* 0 #f) (set! *task-db* #f))))) |
︙ | |||
1206 1207 1208 1209 1210 1211 1212 | 1217 1218 1219 1220 1221 1222 1223 1224 1225 1226 1227 1228 1229 1230 1231 | - + | (common:set-last-run-version))) ;; called in megatest.scm, host-port is string hostname:port ;; ;; NOTE: This is NOT called directly from clients as not all transports support a client running ;; in the same process as the server. ;; |
︙ | |||
1232 1233 1234 1235 1236 1237 1238 | 1243 1244 1245 1246 1247 1248 1249 1250 1251 1252 1253 1254 1255 1256 1257 | - + | (begin (if host-port-in (debug:print 0 *default-log-port* "ERROR: bad host:port")) (if do-exit (exit 1)) #f) (let* ((iface (car host-port)) (port (cadr host-port)) |
︙ | |||
1262 1263 1264 1265 1266 1267 1268 | 1273 1274 1275 1276 1277 1278 1279 1280 1281 1282 1283 1284 1285 1286 1287 1288 1289 1290 1291 1292 1293 1294 1295 1296 1297 1298 1299 | - + - + | ((NOREPLY) #f) ((LOGIN_OK) #t) (else #f)) (loop (read-line) inl)))))) ;; ping the given server ;; |
︙ | |||
1307 1308 1309 1310 1311 1312 1313 | 1318 1319 1320 1321 1322 1323 1324 1325 1326 1327 1328 1329 1330 1331 1332 | - + | #t #f))) ;; Given a run id start a server process ### NOTE ### > file 2>&1 ;; if the run-id is zero and the target-host is set ;; try running on that host ;; incidental: rotate logs in logs/ dir. ;; |
︙ | |||
1351 1352 1353 1354 1355 1356 1357 | 1362 1363 1364 1365 1366 1367 1368 1369 1370 1371 1372 1373 1374 1375 1376 1377 1378 1379 1380 1381 1382 1383 1384 1385 1386 1387 1388 1389 1390 1391 1392 1393 1394 1395 1396 1397 1398 | - + - + - + | (unsetenv "TARGETHOST_LOGF") (if (get-environment-variable "TARGETHOST")(unsetenv "TARGETHOST")) (thread-join! log-rotate) (pop-directory))) ;; kind start up of servers, wait 40 seconds before allowing another server for a given ;; run-id to be launched |
︙ | |||
2440 2441 2442 2443 2444 2445 2446 2447 2448 | 2451 2452 2453 2454 2455 2456 2457 2458 2459 2460 2461 2462 2463 2464 2465 2466 2467 2468 2469 2470 2471 2472 2473 2474 2475 2476 2477 2478 2479 2480 2481 2482 2483 2484 2485 2486 2487 2488 2489 2490 2491 2492 2493 2494 2495 2496 2497 2498 2499 2500 2501 2502 2503 2504 2505 2506 2507 2508 2509 2510 2511 2512 2513 2514 2515 2516 2517 2518 2519 2520 2521 2522 2523 2524 2525 2526 2527 2528 2529 2530 2531 2532 2533 2534 2535 2536 2537 2538 2539 2540 2541 2542 2543 2544 2545 2546 2547 2548 2549 2550 2551 2552 2553 2554 2555 2556 2557 2558 2559 2560 2561 2562 2563 2564 2565 2566 2567 2568 2569 2570 2571 2572 2573 2574 2575 2576 2577 2578 2579 2580 2581 2582 2583 2584 2585 2586 2587 2588 2589 2590 2591 2592 2593 2594 2595 2596 2597 2598 2599 2600 2601 2602 2603 2604 2605 2606 2607 2608 2609 2610 2611 2612 2613 2614 2615 2616 2617 2618 2619 2620 2621 2622 2623 2624 2625 2626 2627 2628 2629 2630 2631 2632 2633 2634 2635 2636 2637 2638 2639 2640 2641 2642 2643 2644 2645 2646 2647 2648 2649 2650 2651 2652 2653 2654 2655 2656 2657 2658 2659 2660 2661 2662 2663 2664 2665 2666 2667 2668 2669 2670 2671 2672 2673 2674 2675 2676 2677 2678 2679 2680 2681 2682 2683 2684 2685 2686 2687 2688 2689 2690 2691 2692 2693 2694 2695 2696 2697 2698 2699 2700 2701 2702 2703 2704 2705 2706 2707 2708 2709 2710 2711 2712 2713 2714 2715 2716 2717 2718 2719 2720 2721 2722 2723 2724 2725 2726 2727 2728 2729 2730 2731 2732 2733 2734 2735 2736 2737 2738 2739 2740 2741 2742 2743 2744 2745 2746 2747 2748 2749 2750 2751 2752 2753 2754 2755 2756 2757 2758 2759 2760 2761 2762 2763 2764 2765 2766 2767 2768 2769 2770 2771 2772 2773 2774 2775 2776 2777 2778 2779 2780 2781 2782 2783 2784 2785 2786 2787 2788 2789 2790 2791 2792 2793 2794 2795 2796 2797 2798 2799 2800 2801 2802 2803 2804 2805 2806 2807 2808 2809 2810 2811 2812 2813 2814 2815 2816 2817 2818 2819 2820 2821 2822 2823 2824 2825 2826 2827 2828 2829 2830 2831 2832 2833 2834 2835 2836 2837 2838 2839 2840 2841 2842 2843 2844 2845 2846 2847 2848 2849 2850 2851 2852 2853 2854 2855 2856 2857 2858 2859 2860 2861 2862 2863 2864 2865 2866 2867 2868 2869 2870 2871 2872 2873 2874 2875 2876 2877 2878 2879 2880 2881 2882 2883 2884 2885 2886 2887 2888 2889 2890 2891 2892 2893 2894 2895 2896 2897 2898 2899 2900 2901 2902 2903 2904 2905 2906 2907 2908 2909 2910 2911 2912 2913 2914 2915 2916 2917 2918 2919 2920 2921 2922 2923 2924 2925 2926 2927 2928 2929 2930 2931 2932 2933 2934 2935 2936 2937 2938 2939 2940 2941 2942 2943 2944 2945 2946 2947 2948 2949 2950 2951 2952 2953 2954 2955 2956 2957 2958 2959 2960 2961 2962 2963 2964 2965 2966 2967 2968 2969 2970 2971 2972 2973 2974 2975 2976 2977 2978 2979 2980 2981 2982 2983 2984 2985 2986 2987 2988 2989 2990 2991 2992 2993 2994 2995 2996 2997 2998 2999 3000 3001 3002 3003 3004 3005 3006 3007 3008 3009 3010 3011 3012 3013 3014 3015 3016 3017 3018 3019 3020 3021 3022 3023 3024 3025 3026 3027 3028 3029 3030 3031 3032 3033 3034 3035 3036 3037 3038 3039 3040 3041 3042 3043 3044 3045 3046 3047 3048 3049 3050 3051 3052 3053 3054 3055 3056 3057 3058 3059 3060 3061 3062 3063 3064 3065 3066 3067 3068 3069 3070 3071 3072 3073 3074 3075 3076 3077 3078 3079 3080 3081 3082 3083 3084 3085 3086 3087 3088 3089 3090 3091 3092 3093 3094 3095 3096 3097 3098 3099 3100 3101 3102 3103 3104 3105 3106 3107 3108 3109 3110 3111 3112 3113 3114 3115 3116 3117 3118 3119 3120 3121 3122 3123 3124 3125 3126 3127 3128 3129 3130 3131 3132 3133 3134 3135 3136 3137 3138 3139 3140 3141 3142 3143 3144 3145 3146 3147 3148 3149 3150 3151 3152 3153 3154 3155 3156 3157 3158 3159 3160 3161 3162 3163 3164 3165 3166 3167 3168 3169 3170 3171 3172 3173 3174 3175 3176 3177 3178 3179 3180 3181 3182 3183 3184 3185 3186 3187 3188 3189 3190 3191 3192 3193 3194 3195 3196 3197 3198 3199 3200 3201 3202 3203 3204 3205 3206 3207 3208 3209 3210 3211 3212 3213 3214 3215 3216 3217 3218 3219 3220 3221 3222 3223 3224 3225 3226 3227 3228 3229 3230 3231 3232 3233 3234 3235 3236 3237 3238 3239 3240 3241 3242 3243 3244 3245 3246 3247 3248 3249 3250 3251 3252 3253 3254 3255 3256 3257 3258 3259 3260 3261 3262 3263 3264 3265 3266 3267 3268 3269 3270 3271 3272 3273 3274 3275 3276 3277 3278 3279 3280 3281 3282 3283 3284 3285 3286 3287 3288 3289 3290 3291 3292 3293 3294 3295 3296 3297 3298 3299 3300 3301 3302 3303 3304 3305 3306 3307 3308 3309 3310 3311 3312 3313 3314 3315 3316 3317 3318 3319 3320 3321 3322 3323 3324 3325 3326 3327 3328 3329 3330 3331 3332 3333 3334 3335 3336 3337 3338 3339 3340 3341 3342 3343 3344 3345 3346 3347 3348 3349 3350 3351 3352 3353 3354 3355 3356 3357 3358 3359 3360 3361 3362 3363 3364 3365 3366 3367 3368 3369 3370 3371 3372 3373 3374 3375 3376 3377 3378 3379 3380 3381 3382 3383 3384 3385 3386 3387 3388 3389 3390 3391 3392 3393 3394 3395 3396 3397 3398 3399 3400 3401 3402 3403 3404 3405 3406 3407 3408 3409 3410 3411 3412 3413 3414 3415 3416 3417 3418 3419 3420 3421 3422 3423 3424 3425 3426 3427 3428 3429 3430 3431 3432 3433 3434 3435 3436 3437 3438 3439 3440 3441 3442 3443 3444 3445 3446 3447 3448 3449 3450 3451 3452 3453 3454 3455 3456 3457 3458 3459 3460 3461 3462 3463 3464 3465 3466 3467 3468 3469 3470 3471 3472 3473 3474 3475 3476 3477 3478 3479 3480 3481 3482 3483 3484 3485 3486 3487 3488 3489 3490 3491 3492 3493 3494 3495 3496 3497 3498 3499 3500 3501 3502 3503 3504 3505 3506 3507 3508 3509 3510 3511 3512 3513 3514 3515 3516 3517 3518 3519 3520 3521 3522 3523 3524 3525 3526 3527 3528 3529 3530 3531 3532 3533 3534 3535 3536 3537 3538 3539 3540 3541 3542 3543 3544 3545 3546 3547 3548 3549 3550 3551 3552 3553 3554 3555 3556 3557 3558 3559 3560 3561 3562 3563 3564 3565 3566 3567 3568 3569 3570 3571 3572 3573 3574 3575 3576 3577 3578 3579 3580 3581 3582 3583 3584 3585 3586 3587 3588 3589 3590 3591 3592 3593 3594 3595 3596 3597 3598 3599 3600 3601 3602 3603 3604 3605 3606 3607 3608 3609 3610 3611 3612 3613 3614 3615 3616 3617 3618 3619 3620 3621 3622 3623 3624 3625 3626 3627 3628 3629 3630 3631 3632 3633 3634 3635 3636 3637 3638 3639 3640 3641 3642 3643 3644 3645 3646 3647 3648 3649 3650 3651 3652 3653 3654 3655 3656 3657 3658 3659 3660 3661 3662 3663 3664 3665 3666 3667 3668 3669 3670 3671 3672 3673 3674 3675 3676 3677 3678 3679 3680 3681 3682 3683 3684 3685 3686 3687 3688 3689 3690 3691 3692 3693 3694 3695 3696 3697 3698 3699 3700 3701 3702 3703 3704 3705 3706 3707 3708 3709 3710 3711 3712 3713 3714 3715 3716 3717 3718 3719 3720 3721 3722 3723 3724 3725 3726 3727 3728 3729 3730 3731 3732 3733 3734 3735 3736 3737 3738 3739 3740 3741 3742 3743 3744 3745 3746 3747 3748 3749 3750 3751 3752 3753 3754 3755 3756 3757 3758 3759 3760 3761 3762 3763 3764 3765 3766 3767 3768 3769 3770 3771 3772 3773 3774 3775 3776 3777 3778 3779 3780 3781 3782 3783 3784 3785 3786 3787 3788 3789 3790 3791 3792 3793 3794 3795 3796 3797 3798 3799 3800 3801 3802 3803 3804 3805 3806 3807 3808 3809 3810 3811 3812 3813 3814 3815 3816 3817 3818 3819 3820 3821 3822 3823 3824 3825 3826 3827 3828 3829 3830 3831 3832 3833 3834 3835 3836 3837 3838 3839 3840 3841 3842 3843 3844 3845 3846 3847 3848 3849 3850 3851 3852 3853 3854 3855 3856 3857 3858 3859 3860 3861 3862 3863 3864 3865 3866 3867 3868 3869 3870 3871 3872 3873 3874 3875 3876 3877 3878 3879 3880 3881 3882 3883 3884 3885 3886 3887 3888 3889 3890 3891 3892 3893 3894 3895 3896 3897 3898 3899 3900 3901 3902 3903 3904 3905 3906 3907 3908 3909 3910 3911 3912 3913 3914 3915 3916 3917 3918 3919 3920 3921 3922 3923 3924 3925 3926 3927 3928 3929 3930 3931 3932 3933 3934 3935 3936 3937 3938 3939 3940 3941 3942 3943 3944 3945 3946 3947 3948 3949 3950 3951 3952 3953 3954 3955 3956 3957 3958 3959 3960 3961 3962 3963 3964 3965 3966 3967 3968 3969 3970 3971 3972 3973 3974 3975 3976 3977 3978 3979 3980 3981 3982 3983 3984 3985 3986 3987 3988 3989 3990 3991 3992 3993 3994 3995 3996 3997 3998 3999 4000 4001 4002 4003 4004 4005 4006 4007 4008 4009 4010 4011 4012 4013 4014 4015 4016 4017 4018 4019 4020 4021 4022 4023 4024 4025 4026 4027 4028 4029 4030 4031 4032 4033 4034 4035 4036 4037 4038 4039 4040 4041 4042 4043 4044 4045 4046 4047 4048 4049 4050 4051 4052 4053 4054 4055 4056 4057 4058 4059 4060 4061 4062 4063 4064 4065 4066 4067 4068 4069 4070 4071 4072 4073 4074 4075 4076 4077 4078 4079 4080 4081 4082 4083 4084 4085 4086 4087 4088 4089 4090 4091 4092 4093 4094 4095 4096 4097 4098 4099 4100 4101 4102 4103 4104 4105 4106 4107 4108 4109 4110 4111 4112 4113 4114 4115 4116 4117 4118 4119 4120 4121 4122 4123 4124 4125 4126 4127 4128 4129 4130 4131 4132 4133 4134 4135 4136 4137 4138 4139 4140 4141 4142 4143 4144 4145 4146 4147 4148 4149 4150 4151 4152 4153 4154 4155 4156 4157 4158 4159 4160 4161 4162 4163 4164 4165 4166 4167 4168 4169 4170 4171 4172 4173 4174 4175 4176 4177 4178 4179 4180 4181 4182 4183 4184 4185 4186 4187 4188 4189 4190 4191 4192 4193 4194 4195 4196 4197 4198 4199 4200 4201 4202 4203 4204 4205 4206 4207 4208 4209 4210 4211 4212 4213 4214 4215 4216 4217 4218 4219 4220 4221 4222 4223 4224 4225 4226 4227 4228 4229 4230 4231 4232 4233 4234 4235 4236 4237 4238 4239 4240 4241 4242 4243 4244 4245 4246 4247 4248 4249 4250 4251 4252 4253 4254 4255 4256 4257 4258 4259 4260 4261 4262 4263 4264 4265 4266 4267 4268 4269 4270 4271 4272 4273 4274 4275 4276 4277 4278 4279 4280 4281 4282 4283 4284 4285 4286 4287 4288 4289 4290 4291 4292 4293 4294 4295 4296 4297 | + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + | (debug:print 0 *default-log-port* " message: " ((condition-property-accessor 'exn 'message) exn)) (debug:print 5 *default-log-port* "exn=" (condition->list exn)) (debug:print 0 *default-log-port* " status: " ((condition-property-accessor 'sqlite3 'status) exn)) (print-call-chain (current-error-port)))) (tests:update-testdat-meta-info db test-id work-area cpuload diskfree minutes) ))) ;;====================================================================== ;; launch a task - this runs on the originating host, tests themselves ;; ;;====================================================================== ;;====================================================================== ;; ezsteps ;;====================================================================== ;; ezsteps were going to be coded as ;; stepname[,predstep1,predstep2 ...] [{VAR1=first,second,third}] command to execute ;; BUT ;; now are ;; stepname {VAR=first,second,third ...} command ... ;; where the {VAR=first,second,third ...} is optional. ;; given an exit code and whether or not logpro was used calculate OK/BAD ;; return #t if we are ok, #f otherwise (define (steprun-good? logpro exitcode) (or (eq? exitcode 0) (and logpro (eq? exitcode 2)))) ;; if handed a string, process it, else look for MT_CMDINFO (define (launch:get-cmdinfo-assoc-list #!key (encoded-cmd #f)) (let ((enccmd (if encoded-cmd encoded-cmd (getenv "MT_CMDINFO")))) (if enccmd (common:read-encoded-string enccmd) '()))) ;; return (conc status ": " comment) from the final section so that ;; the comment can be set in the step record in launch.scm ;; (define (launch:load-logpro-dat run-id test-id stepname) (let ((cname (conc stepname ".dat"))) (if (common:file-exists? cname) (let* ((dat (configf:read-config cname #f #f)) (csvr (db:logpro-dat->csv dat stepname)) (csvt (let-values (((fmt-cell fmt-record fmt-csv) (make-format ","))) (fmt-csv (map list->csv-record csvr)))) (status (configf:lookup dat "final" "exit-status")) (msg (configf:lookup dat "final" "message"))) (if csvt ;; this if blocked stack dump caused by .dat file from logpro being 0-byte. fixed by upgrading logpro (rmt:csv->test-data run-id test-id csvt) (debug:print 0 *default-log-port* "ERROR: no csvdat exists for run-id: " run-id " test-id: " test-id " stepname: " stepname ", check that logpro version is 1.15 or newer")) ;; (debug:print-info 13 *default-log-port* "Error: run-id/test-id/stepname="run-id"/"test-id"/"stepname" => bad csvr="csvr) ;; ) (cond ((equal? status "PASS") "PASS") ;; skip the message part if status is pass (status (conc (configf:lookup dat "final" "exit-status") ": " (if msg msg "no message"))) (else #f))) #f))) (define (launch:runstep ezstep run-id test-id exit-info m tal testconfig) ;;; TODO: deprecate me in favor of ezsteps.scm (let* ((stepname (car ezstep)) ;; do stuff to run the step (stepinfo (cadr ezstep)) ;; (let ((info (cadr ezstep))) ;; (if (proc? info) "" info))) ;; (stepproc (let ((info (cadr ezstep))) ;; (if (proc? info) info #f))) (stepparts (string-match (regexp "^(\\{([^\\}\\{]*)\\}\\s*|)(.*)$") stepinfo)) (stepparams (list-ref stepparts 2)) ;; for future use, {VAR=1,2,3}, run step for each (paramparts (if (string? stepparams) (map (lambda (x)(string-split x "=")) (string-split-fields "[^;]*=[^;]*" stepparams)) '())) (subrun (alist-ref "subrun" paramparts equal?)) (stepcmd (list-ref stepparts 3)) (script "") ; "#!/bin/bash\n") ;; yep, we depend on bin/bash FIXME!!!\ (logpro-file (conc stepname ".logpro")) (html-file (conc stepname ".html")) (dat-file (conc stepname ".dat")) (tconfig-logpro (configf:lookup testconfig "logpro" stepname)) (logpro-used (common:file-exists? logpro-file))) (debug:print 0 *default-log-port* "stepparts: " stepparts ", stepparams: " stepparams ", paramparts: " paramparts ", subrun: " subrun ", stepcmd: " stepcmd) (if (and tconfig-logpro (not logpro-used)) ;; no logpro file found but have a defn in the testconfig (begin (with-output-to-file logpro-file (lambda () (print ";; logpro file extracted from testconfig\n" ";;") (print tconfig-logpro))) (set! logpro-used #t))) ;; NB// can safely assume we are in test-area directory (debug:print 4 *default-log-port* "ezsteps:\n stepname: " stepname " stepinfo: " stepinfo " stepparts: " stepparts " stepparams: " stepparams " stepcmd: " stepcmd) ;; ;; first source the previous environment ;; (let ((prev-env (conc ".ezsteps/" prevstep (if (string-search (regexp "csh") ;; (get-environment-variable "SHELL")) ".csh" ".sh")))) ;; (if (and prevstep (common:file-exists? prev-env)) ;; (set! script (conc script "source " prev-env)))) ;; call the command using mt_ezstep ;; (set! script (conc "mt_ezstep " stepname " " (if prevstep prevstep "x") " " stepcmd)) (debug:print 4 *default-log-port* "script: " script) (rmt:teststep-set-status! run-id test-id stepname "start" "-" #f #f) ;; now launch the actual process (call-with-environment-variables (list (cons "PATH" (conc (get-environment-variable "PATH") ":."))) (lambda () ;; (process-run "/bin/bash" "-c" "exec ls -l /tmp/foobar > /tmp/delme-more.log 2>&1") (let* ((cmd (conc stepcmd " > " stepname ".log 2>&1")) ;; >outfile 2>&1 (pid #f)) (let ((proc (lambda () (set! pid (process-run "/bin/bash" (list "-c" cmd)))))) (if subrun (begin (debug:print-info 0 *default-log-port* "Running without MT_.* environment variables.") (common:without-vars proc "^MT_.*")) (proc))) (with-output-to-file "Makefile.ezsteps" (lambda () (print stepname ".log :") (print "\t" cmd) (if (common:file-exists? (conc stepname ".logpro")) (print "\tlogpro " stepname ".logpro " stepname ".html < " stepname ".log")) (print) (print stepname " : " stepname ".log") (print)) #:append) (rmt:test-set-top-process-pid run-id test-id pid) (let processloop ((i 0)) (let-values (((pid-val exit-status exit-code)(process-wait pid #t))) (mutex-lock! m) (launch:einf-pid-set! exit-info pid) ;; (vector-set! exit-info 0 pid) (launch:einf-exit-status-set! exit-info exit-status) ;; (vector-set! exit-info 1 exit-status) (launch:einf-exit-code-set! exit-info exit-code) ;; (vector-set! exit-info 2 exit-code) (mutex-unlock! m) (if (eq? pid-val 0) (begin (thread-sleep! 2) (processloop (+ i 1)))) ))))) (debug:print-info 0 *default-log-port* "step " stepname " completed with exit code " (launch:einf-exit-code exit-info)) ;; (vector-ref exit-info 2)) ;; now run logpro if needed (if logpro-used (let* ((logpro-exe (or (getenv "LOGPRO_EXE") "logpro")) (pid (process-run (conc "/bin/sh -c '"logpro-exe" "logpro-file " " (conc stepname ".html") " < " stepname ".log > /dev/null'")))) (let processloop ((i 0)) (let-values (((pid-val exit-status exit-code)(process-wait pid #t))) (mutex-lock! m) ;; (make-launch:einf pid: pid exit-status: exit-status exit-code: exit-code) (launch:einf-pid-set! exit-info pid) ;; (vector-set! exit-info 0 pid) (launch:einf-exit-status-set! exit-info exit-status) ;; (vector-set! exit-info 1 exit-status) (launch:einf-exit-code-set! exit-info exit-code) ;; (vector-set! exit-info 2 exit-code) (mutex-unlock! m) (if (eq? pid-val 0) (begin (thread-sleep! 2) (processloop (+ i 1))))) (debug:print-info 0 *default-log-port* "logpro for step " stepname " exited with code " (launch:einf-exit-code exit-info))))) ;; (vector-ref exit-info 2))))) (let ((exinfo (launch:einf-exit-code exit-info)) ;; (vector-ref exit-info 2)) (logfna (if logpro-used (conc stepname ".html") "")) (comment #f)) (if logpro-used (let ((datfile (conc stepname ".dat"))) ;; load the .dat file into the test_data table if it exists (if (common:file-exists? datfile) (set! comment (launch:load-logpro-dat run-id test-id stepname))) (rmt:test-set-log! run-id test-id (conc stepname ".html")))) (rmt:teststep-set-status! run-id test-id stepname "end" exinfo comment logfna)) ;; set the test final status (let* ((process-exit-status (launch:einf-exit-code exit-info)) ;; (vector-ref exit-info 2)) (this-step-status (cond ((and (eq? process-exit-status 2) logpro-used) 'warn) ;; logpro 2 = warnings ((and (eq? process-exit-status 3) logpro-used) 'check) ;; logpro 3 = check ((and (eq? process-exit-status 4) logpro-used) 'waived) ;; logpro 4 = waived ((and (eq? process-exit-status 5) logpro-used) 'abort) ;; logpro 5 = abort ((and (eq? process-exit-status 6) logpro-used) 'skip) ;; logpro 6 = skip ((eq? process-exit-status 0) 'pass) ;; logpro 0 = pass (else 'fail))) (overall-status (cond ((eq? (launch:einf-rollup-status exit-info) 2) 'warn) ;; rollup-status (vector-ref exit-info 3) ((eq? (launch:einf-rollup-status exit-info) 0) 'pass) ;; (vector-ref exit-info 3) (else 'fail))) (next-status (cond ((eq? overall-status 'pass) this-step-status) ((eq? overall-status 'warn) (if (eq? this-step-status 'fail) 'fail 'warn)) ((eq? overall-status 'abort) 'abort) (else 'fail))) (next-state ;; "RUNNING") ;; WHY WAS THIS CHANGED TO NOT USE (null? tal) ?? (cond ((null? tal) ;; more to run? "COMPLETED") (else "RUNNING")))) (debug:print 4 *default-log-port* "Exit value received: " (launch:einf-exit-code exit-info) " logpro-used: " logpro-used " this-step-status: " this-step-status " overall-status: " overall-status " next-status: " next-status " rollup-status: " (launch:einf-rollup-status exit-info)) ;; (vector-ref exit-info 3)) (case next-status ((warn) (launch:einf-rollup-status-set! exit-info 2) ;; (vector-set! exit-info 3 2) ;; rollup-status ;; NB// test-set-status! does rdb calls under the hood (tests:test-set-status! run-id test-id next-state "WARN" (if (eq? this-step-status 'warn) "Logpro warning found" #f) #f)) ((check) (launch:einf-rollup-status-set! exit-info 3) ;; (vector-set! exit-info 3 3) ;; rollup-status ;; NB// test-set-status! does rdb calls under the hood (tests:test-set-status! run-id test-id next-state "CHECK" (if (eq? this-step-status 'check) "Logpro check found" #f) #f)) ((waived) (launch:einf-rollup-status-set! exit-info 4) ;; (vector-set! exit-info 3 3) ;; rollup-status ;; NB// test-set-status! does rdb calls under the hood (tests:test-set-status! run-id test-id next-state "WAIVED" (if (eq? this-step-status 'check) "Logpro waived found" #f) #f)) ((abort) (launch:einf-rollup-status-set! exit-info 5) ;; (vector-set! exit-info 3 4) ;; rollup-status ;; NB// test-set-status! does rdb calls under the hood (tests:test-set-status! run-id test-id next-state "ABORT" (if (eq? this-step-status 'abort) "Logpro abort found" #f) #f)) ((skip) (launch:einf-rollup-status-set! exit-info 6) ;; (vector-set! exit-info 3 4) ;; rollup-status ;; NB// test-set-status! does rdb calls under the hood (tests:test-set-status! run-id test-id next-state "SKIP" (if (eq? this-step-status 'skip) "Logpro skip found" #f) #f)) ((pass) (tests:test-set-status! run-id test-id next-state "PASS" #f #f)) (else ;; 'fail (launch:einf-rollup-status-set! exit-info 1) ;; (vector-set! exit-info 3 1) ;; force fail, this used to be next-state but that doesn't make sense. should always be "COMPLETED" (tests:test-set-status! run-id test-id "COMPLETED" "FAIL" (conc "Failed at step " stepname) #f) ))) logpro-used)) (define (launch:manage-steps run-id test-id item-path fullrunscript ezsteps subrun test-name tconfigreg exit-info m) ;; (let-values ;; (((pid exit-status exit-code) ;; (run-n-wait fullrunscript))) ;; (tests:test-set-status! test-id "RUNNING" "n/a" #f #f) ;; Since we should have a clean slate at this time there is no need to do ;; any of the other stuff that tests:test-set-status! does. Let's just ;; force RUNNING/n/a ;; (thread-sleep! 0.3) ;; (tests:test-force-state-status! run-id test-id "RUNNING" "n/a") (rmt:set-state-status-and-roll-up-items run-id test-name item-path "RUNNING" #f #f) ;; (thread-sleep! 0.3) ;; NFS slowness has caused grief here ;; if there is a runscript do it first (if fullrunscript (let ((pid (process-run fullrunscript))) (rmt:test-set-top-process-pid run-id test-id pid) (let loop ((i 0)) (let-values (((pid-val exit-status exit-code) (process-wait pid #t))) (mutex-lock! m) (launch:einf-pid-set! exit-info pid) ;; (vector-set! exit-info 0 pid) (launch:einf-exit-status-set! exit-info exit-status) ;; (vector-set! exit-info 1 exit-status) (launch:einf-exit-code-set! exit-info exit-code) ;; (vector-set! exit-info 2 exit-code) (launch:einf-rollup-status-set! exit-info exit-code) ;; (vector-set! exit-info 3 exit-code) ;; rollup status (mutex-unlock! m) (if (eq? pid-val 0) (begin (thread-sleep! 2) (loop (+ i 1))) ))))) ;; then, if runscript ran ok (or did not get called) ;; do all the ezsteps (if any) (if (or ezsteps subrun) (let* ((test-run-dir (tests:get-test-path-from-environment)) (testconfig ;; (read-config (conc work-area "/testconfig") #f #t environ-patt: "pre-launch-env-vars")) ;; FIXME??? is allow-system ok here? ;; NOTE: it is tempting to turn off force-create of testconfig but dynamic ;; ezstep names need a full re-eval here. (tests:get-testconfig test-name item-path tconfigreg #t force-create: #t)) ;; 'return-procs))) (ezstepslst (if (hash-table? testconfig) (hash-table-ref/default testconfig "ezsteps" '()) #f))) (if testconfig (hash-table-set! *testconfigs* test-name testconfig) ;; cached for lazy reads later ... (begin (launch:setup) (debug:print 0 *default-log-port* "WARNING: no testconfig found for " test-name " in search path:\n " (string-intersperse (tests:get-tests-search-path *configdat*) "\n ")))) ;; after all that, still no testconfig? Time to abort (if (not testconfig) (begin (debug:print-error 0 *default-log-port* "Failed to resolve megatest.config, runconfigs.config and testconfig issues. Giving up now") (exit 1))) ;; create a proc for the subrun if requested, save that proc in the ezsteps table as the last entry ;; 1. get section [runarun] ;; 2. unset MT_* vars ;; 3. fix target ;; 4. fix runname ;; 5. fix testpatt or calculate it from contour ;; 6. launch the run ;; 7. roll up the run result and or roll up the logpro processed result (when (configf:lookup testconfig "subrun" "runwait") ;; we use runwait as the flag that a subrun is requested (subrun:initialize-toprun-test testconfig test-run-dir) (let* ((mt-cmd (subrun:launch-cmd test-run-dir))) (debug:print-info 0 *default-log-port* "Subrun command is \"" mt-cmd "\"") (set! ezsteps #t) ;; set the needed flag (set! ezstepslst (append (or ezstepslst '()) (list (list "subrun" (conc "{subrun=true} " mt-cmd))))))) ;; process the ezsteps (if ezsteps (begin (if (not (common:file-exists? ".ezsteps"))(create-directory ".ezsteps")) ;; if ezsteps was defined then we are sure to have at least one step but check anyway (if (not (> (length ezstepslst) 0)) (debug:print-error 0 *default-log-port* "ezsteps defined but ezstepslst is zero length") (let loop ((ezstep (car ezstepslst)) (tal (cdr ezstepslst)) (prevstep #f)) (debug:print-info 0 *default-log-port* "Processing ezstep \"" (string-intersperse ezstep " ") "\"") ;; check exit-info (vector-ref exit-info 1) (if (launch:einf-exit-status exit-info) ;; (vector-ref exit-info 1) (let ((logpro-used (launch:runstep ezstep run-id test-id exit-info m tal testconfig)) (stepname (car ezstep))) ;; if logpro-used read in the stepname.dat file (if (and logpro-used (common:file-exists? (conc stepname ".dat"))) (launch:load-logpro-dat run-id test-id stepname)) (if (steprun-good? logpro-used (launch:einf-exit-code exit-info)) (if (not (null? tal)) (loop (car tal) (cdr tal) stepname)) (debug:print 0 *default-log-port* "WARNING: step " (car ezstep) " failed. Stopping"))) (debug:print 0 *default-log-port* "WARNING: a prior step failed, stopping at " ezstep))))))))) (define (launch:monitor-job run-id test-id item-path fullrunscript ezsteps test-name tconfigreg exit-info m work-area runtlim misc-flags) (let* ((update-period (string->number (or (configf:lookup *configdat* "setup" "test-stats-update-period") "30"))) (start-seconds (current-seconds)) (calc-minutes (lambda () (inexact->exact (round (- (current-seconds) start-seconds))))) (kill-tries 0)) ;; (tests:set-full-meta-info #f test-id run-id (calc-minutes) work-area) ;; (tests:set-full-meta-info test-id run-id (calc-minutes) work-area) (tests:set-full-meta-info #f test-id run-id (calc-minutes) work-area 10) (let loop ((minutes (calc-minutes)) (cpu-load (alist-ref 'adj-core-load (common:get-normalized-cpu-load #f))) (disk-free (get-df (current-directory))) (last-sync (current-seconds))) #;(common:telemetry-log "zombie" (conc "launch:monitor-job - top of loop encountered at "(current-seconds)" with last-sync="last-sync)) (let* ((over-time (> (current-seconds) (+ last-sync update-period))) (new-cpu-load (let* ((load (alist-ref 'adj-core-load (common:get-normalized-cpu-load #f))) (delta (abs (- load cpu-load)))) (if (> delta 0.1) ;; don't bother updating with small changes load #f))) (new-disk-free (let* ((df (if over-time ;; only get df every 30 seconds (get-df (current-directory)) disk-free)) (delta (abs (- df disk-free)))) (if (and (> df 0) (> (/ delta df) 0.1)) ;; (> delta 200) ;; ignore changes under 200 Meg df #f))) (do-sync (or new-cpu-load new-disk-free over-time)) (test-info (rmt:get-test-info-by-id run-id test-id)) (state (db:test-get-state test-info)) (status (db:test-get-status test-info)) (kill-reason "no kill reason specified") (kill-job? #f)) #;(common:telemetry-log "zombie" (conc "launch:monitor-job - decision time encountered at "(current-seconds)" with last-sync="last-sync" do-sync="do-sync" over-time="over-time" update-period="update-period)) (cond ((test-get-kill-request run-id test-id) (set! kill-reason "KILLING TEST since received kill request (KILLREQ)") (set! kill-job? #t)) ((and runtlim (> (- (current-seconds) start-seconds) runtlim)) (set! kill-reason (conc "KILLING TEST DUE TO TIME LIMIT EXCEEDED! Runtime=" (- (current-seconds) start-seconds) " seconds, limit=" runtlim)) (set! kill-job? #t)) ((equal? status "DEAD") (tests:update-central-meta-info run-id test-id new-cpu-load new-disk-free (calc-minutes) #f #f) (rmt:set-state-status-and-roll-up-items run-id test-id 'foo "RUNNING" "n/a" "was marked dead; really still running.") ;;(set! kill-reason "KILLING TEST because it was marked as DEAD by launch:handle-zombie-tests (might indicate really overloaded server or else overzealous setup.deadtime)") ;; MARK RUNNING (set! kill-job? #f))) (debug:print 4 *default-log-port* "cpu: " new-cpu-load " disk: " new-disk-free " last-sync: " last-sync " do-sync: " do-sync) (launch:handle-zombie-tests run-id) (when do-sync ;;(with-output-to-file (conc (getenv "MT_TEST_RUN_DIR") "/last-loadinfo.log" #:append) ;; (lambda () (pp (list (current-seconds) new-cpu-load new-disk-free (calc-minutes))))) #;(common:telemetry-log "zombie" (conc "launch:monitor-job - dosync started at "(current-seconds))) (tests:update-central-meta-info run-id test-id new-cpu-load new-disk-free (calc-minutes) #f #f) #;(common:telemetry-log "zombie" (conc "launch:monitor-job - dosync finished at "(current-seconds)))) (if kill-job? (begin (debug:print-info 0 *default-log-port* "proceeding to kill test: "kill-reason) (mutex-lock! m) ;; NOTE: The pid can change as different steps are run. Do we need handshaking between this ;; section and the runit section? Or add a loop that tries three times with a 1/4 second ;; between tries? (let* ((pid1 (launch:einf-pid exit-info)) ;; (vector-ref exit-info 0)) (pid2 (rmt:test-get-top-process-pid run-id test-id)) (pids (delete-duplicates (filter number? (list pid1 pid2))))) (if (not (null? pids)) (begin (for-each (lambda (pid) (handle-exceptions exn (begin (debug:print-info 0 *default-log-port* "Unable to kill process with pid " pid ", possibly already killed.") (debug:print 0 *default-log-port* " message: " ((condition-property-accessor 'exn 'message) exn))) (debug:print 0 *default-log-port* "WARNING: Request received to kill job " pid) ;; " (attempt # " kill-tries ")") (debug:print-info 0 *default-log-port* "Signal mask=" (signal-mask)) ;; (if (process:alive? pid) ;; (begin (map (lambda (pid-num) (process-signal pid-num signal/term)) (process:get-sub-pids pid)) (thread-sleep! 5) ;; (if (process:process-alive? pid) (map (lambda (pid-num) (handle-exceptions exn #f (process-signal pid-num signal/kill))) (process:get-sub-pids pid)))) ;; (debug:print-info 0 *default-log-port* "not killing process " pid " as it is not alive")))) pids) ;; BB: question to Matt -- does the tests:test-state-status! encompass rollup to toplevel? If not, should it? (tests:test-set-status! run-id test-id "KILLED" "KILLED" (conc (args:get-arg "-m")" "kill-reason) #f)) ;; BB ADDED kill-reason -- confirm OK with Matt (begin (debug:print-error 0 *default-log-port* "Nothing to kill, pid1=" pid1 ", pid2=" pid2) (tests:test-set-status! run-id test-id "KILLED" "FAILED TO KILL" (conc (args:get-arg "-m")" "kill-reason) #f) ;; BB ADDED kill-reason -- confirm OK with Matt ))) (mutex-unlock! m) ;; no point in sticking around. Exit now. But run end of run before exiting? (launch:end-of-run-check run-id) (exit))) (if (hash-table-ref/default misc-flags 'keep-going #f) (begin (thread-sleep! 3) ;; (+ 3 (random 6))) ;; add some jitter to the call home time to spread out the db accesses (if (hash-table-ref/default misc-flags 'keep-going #f) ;; keep originals for cpu-load and disk-free unless they change more than the allowed delta (loop (calc-minutes) (or new-cpu-load cpu-load) (or new-disk-free disk-free) (if do-sync (current-seconds) last-sync))))))) (tests:update-central-meta-info run-id test-id (get-cpu-load) (get-df (current-directory))(calc-minutes) #f #f))) ;; NOTE: Checking twice for keep-going is intentional ;; set up needed environment variables given a run-id and optionally a target, itempath etc. ;; (define (runs:set-megatest-env-vars run-id #!key (inkeys #f)(inrunname #f)(inkeyvals #f)(intarget #f)(testname #f)(itempath #f)) ;;(bb-check-path msg: "runs:set-megatest-env-vars entry") (let* ((target (or intarget (common:args-get-target) (get-environment-variable "MT_TARGET"))) (keys (if inkeys inkeys (rmt:get-keys))) (keyvals (if inkeyvals inkeyvals (keys:target->keyval keys target))) (vals (hash-table-ref/default *env-vars-by-run-id* run-id #f)) (link-tree (common:get-linktree))) ;; (configf:lookup *configdat* "setup" "linktree"))) (if testname (setenv "MT_TEST_NAME" testname)) (if itempath (setenv "MT_ITEMPATH" itempath)) ;; get the info from the db and put it in the cache (if link-tree (setenv "MT_LINKTREE" link-tree) (debug:print-error 0 *default-log-port* "linktree not set, should be set in megatest.config in [setup] section.")) (if (not vals) (let ((ht (make-hash-table))) (hash-table-set! *env-vars-by-run-id* run-id ht) (set! vals ht) (for-each (lambda (key) (hash-table-set! vals (car key) (cadr key))) keyvals))) ;; from the cached data set the vars (hash-table-for-each vals (lambda (key val) (debug:print 2 *default-log-port* "setenv " key " " val) (safe-setenv key val))) ;;(bb-check-path msg: "runs:set-megatest-env-vars block 1") ;;(BB> "*env-vars-by-run-id*/runid("run-id" vals="(hash-table->alist vals)) (if (not (get-environment-variable "MT_TARGET"))(setenv "MT_TARGET" target)) ;; we had a case where there was an exception generated by the hash-table-ref ;; due to *configdat* being #f Adding a handle and exit (let fatal-loop ((count 0)) (handle-exceptions exn (let ((call-chain (get-call-chain)) (msg ((condition-property-accessor 'exn 'message) exn))) (if (< count 5) (begin ;; this call is colliding, do some crude stuff to fix it. (debug:print 0 *default-log-port* "ERROR: *configdat* was inaccessible! This should never happen. Retry #" count) (launch:setup force-reread: #t) (fatal-loop (+ count 1))) (begin (debug:print 0 *default-log-port* "FATAL: *configdat* was inaccessible! This should never happen. Retried " count " times. Message: " msg) (debug:print 0 *default-log-port* "Call chain:") (with-output-to-port *default-log-port* (lambda () (print "*configdat* is >>"*configdat*"<<") (pp *configdat*) (pp call-chain))) (exit 1)))) ;;(bb-check-path msg: "runs:set-megatest-env-vars block 1.5") (when (or (not *configdat*) (not (hash-table? *configdat*))) (debug:print 0 *default-log-port* "WARNING: *configdat* was inaccessible! This should never happen. Brute force reread.") ;;(BB> "ERROR: *configdat* was inaccessible! This should never happen. Brute force reread.") (thread-sleep! 2) ;; assuming nfs lag. (launch:setup force-reread: #t)) (alist->env-vars (hash-table-ref/default *configdat* "env-override" '())))) ;;;; environment is tainted HERE in this let block. ;;(bb-check-path msg: "runs:set-megatest-env-vars block 2") ;; Lets use this as an opportunity to put MT_RUNNAME in the environment (let ((runname (if inrunname inrunname (rmt:get-run-name-from-id run-id)))) (if runname (setenv "MT_RUNNAME" runname) (debug:print-error 0 *default-log-port* "no value for runname for id " run-id))) (setenv "MT_RUN_AREA_HOME" *toppath*) ;; if a testname and itempath are available set the remaining appropriate variables (if testname (setenv "MT_TEST_NAME" testname)) (if itempath (setenv "MT_ITEMPATH" itempath)) ;;(bb-check-path msg: "runs:set-megatest-env-vars block 3") (if (and testname link-tree) (setenv "MT_TEST_RUN_DIR" (conc (getenv "MT_LINKTREE") "/" (getenv "MT_TARGET") "/" (getenv "MT_RUNNAME") "/" (getenv "MT_TEST_NAME") (if (and itempath (not (equal? itempath ""))) (conc "/" itempath) "")))))) (define (launch:execute encoded-cmd) (let* ((cmdinfo (common:read-encoded-string encoded-cmd)) (tconfigreg #f)) (setenv "MT_CMDINFO" encoded-cmd) ;;(bb-check-path msg: "launch:execute incoming") (if (list? cmdinfo) ;; ((testpath /tmp/mrwellan/jazzmind/src/example_run/tests/sqlitespeed) ;; (test-name sqlitespeed) (runscript runscript.rb) (db-host localhost) (run-id 1)) (let* ((testpath (assoc/default 'testpath cmdinfo)) ;; testpath is the test spec area (top-path (assoc/default 'toppath cmdinfo)) (work-area (assoc/default 'work-area cmdinfo)) ;; work-area is the test run area (test-name (assoc/default 'test-name cmdinfo)) (runscript (assoc/default 'runscript cmdinfo)) (ezsteps (assoc/default 'ezsteps cmdinfo)) (subrun (assoc/default 'subrun cmdinfo)) ;; (runremote (assoc/default 'runremote cmdinfo)) ;; (transport (assoc/default 'transport cmdinfo)) ;; not used ;; (serverinf (assoc/default 'serverinf cmdinfo)) ;; (port (assoc/default 'port cmdinfo)) (serverurl (assoc/default 'serverurl cmdinfo)) (homehost (assoc/default 'homehost cmdinfo)) (run-id (assoc/default 'run-id cmdinfo)) (test-id (assoc/default 'test-id cmdinfo)) (target (assoc/default 'target cmdinfo)) (areaname (assoc/default 'areaname cmdinfo)) (itemdat (assoc/default 'itemdat cmdinfo)) (env-ovrd (assoc/default 'env-ovrd cmdinfo)) (set-vars (assoc/default 'set-vars cmdinfo)) ;; pre-overrides from -setvar (runname (assoc/default 'runname cmdinfo)) (megatest (assoc/default 'megatest cmdinfo)) (runtlim (assoc/default 'runtlim cmdinfo)) (contour (assoc/default 'contour cmdinfo)) (item-path (item-list->path itemdat)) (mt-bindir-path (assoc/default 'mt-bindir-path cmdinfo)) (keys #f) (keyvals #f) (fullrunscript (if (not runscript) #f (if (substring-index "/" runscript) runscript ;; use unadultered if contains slashes (let ((fulln (conc work-area "/" runscript))) (if (and (common:file-exists? fulln) (file-execute-access? fulln)) fulln runscript))))) ;; assume it is on the path (check-work-area (lambda () ;; NFS might not have propagated the directory meta data to the run host - give it time if needed (let loop ((count 0)) (if (or (common:directory-exists? work-area) (> count 10)) (change-directory work-area) (begin (debug:print 0 *default-log-port* "INFO: Not starting job yet - directory " work-area " not found") (thread-sleep! 10) (loop (+ count 1))))) (if (not (string=? (common:real-path work-area)(common:real-path (current-directory)))) (begin (debug:print 0 *default-log-port* "INFO: we are expecting to be in directory " work-area "\n" " but we are actually in the directory " (current-directory) "\n" " doing another change dir.") (change-directory work-area))) ;; spot check that the files in testpath are available. Too often NFS delays cause problems here. (let ((files (glob (conc testpath "/*"))) (bad-files '())) (for-each (lambda (fullname) (let* ((fname (pathname-strip-directory fullname)) (targn (conc work-area "/" fname))) (if (not (file-exists? targn)) (set! bad-files (cons fname bad-files))))) files) (if (not (null? bad-files)) (begin (debug:print 0 *default-log-port* "INFO: test data from " testpath " not copied properly or filesystem problems causing data to not be found. Re-running the copy command.") (debug:print 0 *default-log-port* "INFO: missing files from " work-area ": " (string-intersperse bad-files ", ")) (launch:test-copy testpath work-area)))) ;; one more time, change to the work-area directory (change-directory work-area))) ) ;; let* (if contour (setenv "MT_CONTOUR" contour)) ;; immediated set some key variables from CMDINFO data, yes, these will be set again below ... ;; (setenv "MT_TESTSUITENAME" areaname) (setenv "MT_RUN_AREA_HOME" top-path) (set! *toppath* top-path) (change-directory *toppath*) ;; temporarily switch to the run area home (setenv "MT_TEST_RUN_DIR" work-area) (launch:setup) ;; should be properly in the run area home now (if contour (setenv "MT_CONTOUR" contour)) ;; immediated set some key variables from CMDINFO data, yes, these will be set again below ... ;; (setenv "MT_TESTSUITENAME" areaname) (setenv "MT_RUN_AREA_HOME" top-path) (set! *toppath* top-path) (change-directory *toppath*) ;; temporarily switch to the run area home (setenv "MT_TEST_RUN_DIR" work-area) (launch:setup) ;; should be properly in the run area home now (set! tconfigreg (tests:get-all)) ;; mapping of testname => test source path (let ((sighand (lambda (signum) ;; (signal-mask! signum) ;; to mask or not? seems to cause issues in exiting (if (eq? signum signal/stop) (debug:print-error 0 *default-log-port* "attempt to STOP process. Exiting.")) (set! *time-to-exit* #t) (print "Received signal " signum ", cleaning up before exit (set this test to COMPLETED/ABORT) . Please wait...") (let ((th1 (make-thread (lambda () (print "set test to COMPLETED/ABORT begin.") (rmt:test-set-state-status run-id test-id "COMPLETED" "ABORT" "received kill signal") (print "set test to COMPLETED/ABORT complete.") (print "Killed by signal " signum ". Exiting") (exit 1)))) (th2 (make-thread (lambda () (thread-sleep! 20) (debug:print 0 *default-log-port* "Done") (exit 4))))) (thread-start! th2) (thread-start! th1) (thread-join! th2))))) (set-signal-handler! signal/int sighand) (set-signal-handler! signal/term sighand) ) ;; (set-signal-handler! signal/stop sighand) ;; Do not run the test if it is REMOVING, RUNNING, KILLREQ or REMOTEHOSTSTART, ;; Mark the test as REMOTEHOSTSTART *IMMEDIATELY* ;; (let* ((test-info (rmt:get-test-info-by-id run-id test-id)) (test-host (if test-info (db:test-get-host test-info) (begin (debug:print 0 *default-log-port* "ERROR: failed to find a record for test-id " test-id ", exiting.") (exit)))) (test-pid (db:test-get-process_id test-info))) (cond ;; -mrw- I'm removing KILLREQ from this list so that a test in KILLREQ state is treated as a "do not run" flag. ((member (db:test-get-state test-info) '("INCOMPLETE" "KILLED" "UNKNOWN" "STUCK")) ;; prior run of this test didn't complete, go ahead and try to rerun (debug:print 0 *default-log-port* "INFO: test is INCOMPLETE or KILLED, treat this execute call as a rerun request") ;; (tests:test-force-state-status! run-id test-id "REMOTEHOSTSTART" "n/a") (rmt:general-call 'set-test-start-time #f test-id) (rmt:test-set-state-status run-id test-id "REMOTEHOSTSTART" "n/a" #f) ) ;; prime it for running ((member (db:test-get-state test-info) '("RUNNING" "REMOTEHOSTSTART")) (if (process:alive-on-host? test-host test-pid) (debug:print-error 0 *default-log-port* "test state is " (db:test-get-state test-info) " and process " test-pid " is still running on host " test-host ", cannot proceed") (exit))) ((not (member (db:test-get-state test-info) '("REMOVING" "REMOTEHOSTSTART" "RUNNING" "KILLREQ"))) ;; (tests:test-force-state-status! run-id test-id "REMOTEHOSTSTART" "n/a") (rmt:general-call 'set-test-start-time #f test-id) (rmt:test-set-state-status run-id test-id "REMOTEHOSTSTART" "n/a" #f) ) (else ;; (member (db:test-get-state test-info) '("REMOVING" "REMOTEHOSTSTART" "RUNNING" "KILLREQ")) (debug:print-error 0 *default-log-port* "test state is " (db:test-get-state test-info) ", cannot proceed") (exit)))) ;; cleanup prior execution's steps (rmt:delete-steps-for-test! run-id test-id) (debug:print 2 *default-log-port* "Executing " test-name " (id: " test-id ") on " (get-host-name)) (set! keys (rmt:get-keys)) ;; (runs:set-megatest-env-vars run-id inkeys: keys inkeyvals: keyvals) ;; these may be needed by the launching process ;; one of these is defunct/redundant ... (if (not (launch:setup force-reread: #t)) (begin (debug:print 0 *default-log-port* "Failed to setup, exiting") ;; (sqlite3:finalize! db) ;; (sqlite3:finalize! tdb) (exit 1))) ;; validate that the test run area is available (check-work-area) ;; still need to go back to run area home for next couple steps (change-directory *toppath*) ;; NOTE: Current order is to process runconfigs *before* setting the MT_ vars. This ;; seems non-ideal but could well break stuff ;; BUG? BUG? BUG? (let ((rconfig (full-runconfigs-read)) ;; (read-config (conc *toppath* "/runconfigs.config") #f #t sections: (list "default" target)))) (wconfig (configf:read-config "waivers.config" #f #t sections: `( "default" ,target )))) ;; read the waivers config if it exists ;; (setup-env-defaults (conc *toppath* "/runconfigs.config") run-id (make-hash-table) keyvals target) ;; (set-run-config-vars run-id keyvals target) ;; (db:get-target db run-id)) ;; Now have runconfigs data loaded, set environment vars (for-each (lambda (section) (for-each (lambda (varval) (let ((var (car varval)) (val (cadr varval))) (if (and (string? var)(string? val)) (begin (safe-setenv var (configf:eval-string-in-environment val))) ;; val) (debug:print-error 0 *default-log-port* "bad variable spec, " var "=" val)))) (configf:get-section rconfig section))) (list "default" target))) ;;(bb-check-path msg: "launch:execute post block 1") ;; NFS might not have propagated the directory meta data to the run host - give it time if needed (let loop ((count 0)) (if (or (common:file-exists? work-area) (> count 10)) (change-directory work-area) (begin (debug:print 0 *default-log-port* "INFO: Not starting job yet - directory " work-area " not found") (thread-sleep! 10) (loop (+ count 1))))) ;; now we can switch to the work-area? (change-directory work-area) ;;(bb-check-path msg: "launch:execute post block 1.5") ;; (change-directory work-area) (set! keyvals (keys:target->keyval keys target)) ;; apply pre-overrides before other variables. The pre-override vars must not ;; clobbers things from the official sources such as megatest.config and runconfigs.config (if (string? set-vars) (let ((varpairs (string-split set-vars ","))) (debug:print 4 *default-log-port* "varpairs: " varpairs) (map (lambda (varpair) (let ((varval (string-split varpair "="))) (if (eq? (length varval) 2) (let ((var (car varval)) (val (cadr varval))) (debug:print 1 *default-log-port* "Adding pre-var/val " var " = " val " to the environment") (setenv var val))))) varpairs))) ;;(bb-check-path msg: "launch:execute post block 2") (for-each (lambda (varval) (let ((var (car varval)) (val (cadr varval))) (if val (setenv var val) (begin (debug:print-error 0 *default-log-port* "required variable " var " does not have a valid value. Exiting") (exit))))) (list (list "MT_TEST_RUN_DIR" work-area) (list "MT_TEST_NAME" test-name) (list "MT_ITEM_INFO" (conc itemdat)) (list "MT_ITEMPATH" item-path) (list "MT_RUNNAME" runname) (list "MT_MEGATEST" megatest) (list "MT_TARGET" target) (list "MT_LINKTREE" (common:get-linktree)) ;; (configf:lookup *configdat* "setup" "linktree")) (list "MT_TESTSUITENAME" (common:get-testsuite-name)))) ;;(bb-check-path msg: "launch:execute post block 3") (if mt-bindir-path (setenv "PATH" (conc (getenv "PATH") ":" mt-bindir-path))) ;;(bb-check-path msg: "launch:execute post block 4") ;; (change-directory top-path) ;; Can setup as client for server mode now ;; (client:setup) ;; environment overrides are done *before* the remaining critical envars. (alist->env-vars env-ovrd) ;;(bb-check-path msg: "launch:execute post block 41") (runs:set-megatest-env-vars run-id inkeys: keys inkeyvals: keyvals) ;;(bb-check-path msg: "launch:execute post block 42") (set-item-env-vars itemdat) ;;(bb-check-path msg: "launch:execute post block 43") (let ((blacklist (configf:lookup *configdat* "setup" "blacklistvars"))) (if blacklist (let ((vars (string-split blacklist))) (save-environment-as-files "megatest" ignorevars: vars) (for-each (lambda (var) (unsetenv var)) vars)) (save-environment-as-files "megatest"))) ;;(bb-check-path msg: "launch:execute post block 44") ;; open-run-close not needed for test-set-meta-info ;; (tests:set-full-meta-info #f test-id run-id 0 work-area) ;; (tests:set-full-meta-info test-id run-id 0 work-area) (tests:set-full-meta-info #f test-id run-id 0 work-area 10) ;; (thread-sleep! 0.3) ;; NFS slowness has caused grief here (if (args:get-arg "-xterm") (set! fullrunscript "xterm") (if (and fullrunscript (common:file-exists? fullrunscript) (not (file-execute-access? fullrunscript))) (system (conc "chmod ug+x " fullrunscript)))) ;; We are about to actually kick off the test ;; so this is a good place to remove the records for ;; any previous runs ;; (db:test-remove-steps db run-id testname itemdat) ;; now is also a good time to write the .testconfig file (let* ((tconfig-fname (conc work-area "/.testconfig")) (tconfig-tmpfile (conc tconfig-fname ".tmp")) (tconfig (tests:get-testconfig test-name item-path tconfigreg #t force-create: #t))) ;; 'return-procs))) (configf:write-alist tconfig tconfig-tmpfile) (file-move tconfig-tmpfile tconfig-fname #t)) ;; (let* ((m (make-mutex)) (kill-job? #f) (exit-info (make-launch:einf pid: #t exit-status: #t exit-code: #t rollup-status: 0)) ;; pid exit-status exit-code (i.e. process was successfully run) rollup-status (job-thread #f) ;; (keep-going #t) (misc-flags (let ((ht (make-hash-table))) (hash-table-set! ht 'keep-going #t) ht)) (runit (lambda () (launch:manage-steps run-id test-id item-path fullrunscript ezsteps subrun test-name tconfigreg exit-info m))) (monitorjob (lambda () (launch:monitor-job run-id test-id item-path fullrunscript ezsteps test-name tconfigreg exit-info m work-area runtlim misc-flags))) (th1 (make-thread monitorjob "monitor job")) (th2 (make-thread runit "run job"))) (set! job-thread th2) (thread-start! th1) (thread-start! th2) (thread-join! th2) (debug:print-info 0 *default-log-port* "Megatest exectute of test " test-name ", item path " item-path " complete. Notifying the db ...") (hash-table-set! misc-flags 'keep-going #f) (thread-join! th1) (thread-sleep! 1) ;; givbe thread th1 a chance to be done TODO: Verify this is needed. At 0.1 I was getting fail to stop, increased to total of 1.1 sec. (mutex-lock! m) (let* ((item-path (item-list->path itemdat)) ;; only state and status needed - use lazy routine (testinfo (rmt:get-testinfo-state-status run-id test-id))) ;; Am I completed? (if (member (db:test-get-state testinfo) '("REMOTEHOSTSTART" "RUNNING")) ;; NOTE: It should *not* be REMOTEHOSTSTART but for reasons I don't yet understand it sometimes gets stuck in that state ;; (not (equal? (db:test-get-state testinfo) "COMPLETED")) (let ((new-state (if kill-job? "KILLED" "COMPLETED") ;; (if (eq? (vector-ref exit-info 2) 0) ;; exited with "good" status ;; "COMPLETED" ;; (db:test-get-state testinfo))) ;; else preseve the state as set within the test ) (new-status (cond ((not (launch:einf-exit-status exit-info)) "FAIL") ;; job failed to run ... (vector-ref exit-info 1) ((eq? (launch:einf-rollup-status exit-info) 0) ;; (vector-ref exit-info 3) ;; if the current status is AUTO then defer to the calculated value (i.e. leave this AUTO) (if (equal? (db:test-get-status testinfo) "AUTO") "AUTO" "PASS")) ((eq? (launch:einf-rollup-status exit-info) 1) "FAIL") ;; (vector-ref exit-info 3) ((eq? (launch:einf-rollup-status exit-info) 2) ;; (vector-ref exit-info 3) ;; if the current status is AUTO the defer to the calculated value but qualify (i.e. make this AUTO-WARN) (if (equal? (db:test-get-status testinfo) "AUTO") "AUTO-WARN" "WARN")) ((eq? (launch:einf-rollup-status exit-info) 3) "CHECK") ((eq? (launch:einf-rollup-status exit-info) 4) "WAIVED") ((eq? (launch:einf-rollup-status exit-info) 5) "ABORT") ((eq? (launch:einf-rollup-status exit-info) 6) "SKIP") (else "FAIL")))) ;; (db:test-get-status testinfo))) (debug:print-info 1 *default-log-port* "Test exited in state=" (db:test-get-state testinfo) ", setting state/status based on exit code of " (launch:einf-exit-status exit-info) " and rollup-status of " (launch:einf-rollup-status exit-info)) (tests:test-set-status! run-id test-id new-state new-status (args:get-arg "-m") #f) ;; need to update the top test record if PASS or FAIL and this is a subtest ;; NO NEED TO CALL set-state-status-and-roll-up-items HERE, THIS IS DONE IN set-state-status-and-roll-up-items called by tests:test-set-status! )) ;; for automated creation of the rollup html file this is a good place... (if (not (equal? item-path "")) (tests:summarize-items run-id test-id test-name #f)) (tests:summarize-test run-id test-id) ;; don't force - just update if no (rmt:update-run-stats run-id (rmt:get-raw-run-stats run-id))) (mutex-unlock! m) (launch:end-of-run-check run-id ) (debug:print 2 *default-log-port* "Output from running " fullrunscript ", pid " (launch:einf-pid exit-info) " in work area " work-area ":\n====\n exit code " (launch:einf-exit-code exit-info) "\n" "====\n") (if (not (launch:einf-exit-status exit-info)) (exit 4)))) ))) (define (launch:is-test-alive host pid) (if (and host pid (not (equal? host "n/a"))) (let* ((cmd (conc "ssh " host " pstree -A " pid)) (output (with-input-from-pipe cmd read-lines))) (print "cmd: " cmd "\n op: " output ) (if(eq? (length output) 0) #f #t)) #t)) (define (launch:kill-tests-if-dead run-id) (let* ((running-tests (rmt:get-tests-for-run run-id "%" `("RUNNING" "LAUNCHED" "REMOTEHOSTSTART") `() #f #f #f #f #f #f #f #f))) (let loop ((running-test (car running-tests)) (tal (cdr running-tests)) (kill-cnt 0)) (let* ((test-name (vector-ref running-test 2)) (item-path (vector-ref running-test 11)) (test-id (vector-ref running-test 0)) (host (vector-ref running-test 6)) (pid (rmt:test-get-top-process-pid run-id test-id)) (event-time (vector-ref running-test 5)) (duration (vector-ref running-test 12)) (flag 0) (curr-time (current-seconds))) (if (and (< (+ event-time duration 600) curr-time) (not (launch:is-test-alive host pid))) ;;test has not updated duration in last 10 min then likely its not running but confirm before marking it as killed (begin (debug:print 0 *default-log-port* "test " test-name "/" item-path " needs to be killed") (set! flag 1) (rmt:set-state-status-and-roll-up-items run-id test-name item-path "KILLREQ" "n/a" #f))) (if (not (null? tal)) (loop (car tal) (cdr tal) (+ kill-cnt flag)) (+ kill-cnt flag)))))) ;; DO NOT USE - caching of configs is handled in launch:setup now. ;; (define (launch:cache-config) ;; if we have a linktree and -runtests and -target and the directory exists dump the config ;; to megatest-(current-seconds).cfg and symlink it to megatest.cfg (if (and *configdat* (or (args:get-arg "-run") (args:get-arg "-runtests") (args:get-arg "-execute"))) (let* ((linktree (common:get-linktree)) ;; (get-environment-variable "MT_LINKTREE")) (target (common:args-get-target exit-if-bad: #t)) (runname (or (args:get-arg "-runname") (args:get-arg ":runname") (getenv "MT_RUNNAME"))) (fulldir (conc linktree "/" target "/" runname))) (if (and linktree (common:file-exists? linktree)) ;; can't proceed without linktree (begin (debug:print-info 0 *default-log-port* "Have -run with target=" target ", runname=" runname ", fulldir=" fulldir ", testpatt=" (or (args:get-arg "-testpatt") "%")) (if (not (common:file-exists? fulldir)) (create-directory fulldir #t)) ;; need to protect with exception handler (if (and target runname (common:file-exists? fulldir)) (let ((tmpfile (conc fulldir "/.megatest.cfg." (current-seconds))) (targfile (conc fulldir "/.megatest.cfg-" megatest-version "-" megatest-fossil-hash)) (rconfig (conc fulldir "/.runconfig." megatest-version "-" megatest-fossil-hash))) (if (common:file-exists? rconfig) ;; only cache megatest.config AFTER runconfigs has been cached (begin (debug:print-info 0 *default-log-port* "Caching megatest.config in " tmpfile) (if (not (common:in-running-test?)) (configf:write-alist *configdat* tmpfile)) (system (conc "ln -sf " tmpfile " " targfile)))) ))) (debug:print-info 1 *default-log-port* "No linktree yet, no caching configs."))))) ;; gather available information, if legit read configs in this order: ;; ;; if have cache; ;; read it a return it ;; else ;; megatest.config (do not cache) ;; runconfigs.config (cache if all vars avail) ;; megatest.config (cache if all vars avail) ;; returns: ;; *toppath* ;; side effects: ;; sets; *configdat* (megatest.config info) ;; *runconfigdat* (runconfigs.config info) ;; *configstatus* (status of the read data) ;; (define (launch:setup #!key (force-reread #f) (areapath #f)) (mutex-lock! *launch-setup-mutex*) (if (and *toppath* (eq? *configstatus* 'fulldata) (not force-reread)) ;; got it all (begin (debug:print 2 *default-log-port* "NOTE: skipping launch:setup-body call since we have fulldata") (mutex-unlock! *launch-setup-mutex*) *toppath*) (let ((res (launch:setup-body force-reread: force-reread areapath: areapath))) (mutex-unlock! *launch-setup-mutex*) res))) ;; return paths depending on what info is available. ;; (define (launch:get-cache-file-paths areapath toppath target mtconfig) (let* ((use-cache (common:use-cache?)) (runname (common:args-get-runname)) (linktree (common:get-linktree)) (testname (common:get-full-test-name)) (rundir (if (and runname target linktree) (common:directory-writable? (conc linktree "/" target "/" runname)) #f)) (testdir (if (and rundir testname) (common:directory-writable? (conc rundir "/" testname)) #f)) (cachedir (or testdir rundir)) (mtcachef (and cachedir (conc cachedir "/" ".megatest.cfg-" megatest-version "-" megatest-fossil-hash))) (rccachef (and cachedir (conc cachedir "/" ".runconfigs.cfg-" megatest-version "-" megatest-fossil-hash)))) (debug:print-info 6 *default-log-port* "runname=" runname "\n linktree=" linktree "\n testname=" testname "\n rundir=" rundir "\n testdir=" testdir "\n cachedir=" cachedir "\n mtcachef=" mtcachef "\n rccachef=" rccachef) (cons mtcachef rccachef))) (define (launch:setup-body #!key (force-reread #f) (areapath #f)) (if (and (eq? *configstatus* 'fulldata) *toppath* (not force-reread)) ;; no need to reprocess *toppath* ;; return toppath (let* ((use-cache (common:use-cache?)) ;; BB- use-cache checks *configdat* for use-cache setting. We do not have *configdat*. Bootstrapping problem here. (toppath (or *toppath* areapath (getenv "MT_RUN_AREA_HOME"))) ;; preserve toppath (target (common:args-get-target)) (sections (if target (list "default" target) #f)) ;; for runconfigs (mtconfig (or (args:get-arg "-config") "megatest.config")) ;; allow overriding megatest.config (cachefiles (launch:get-cache-file-paths areapath toppath target mtconfig)) ;; checking for null cachefiles should not be necessary, I was seeing error car of '(), might be a chicken bug or a red herring ... (mtcachef (if (null? cachefiles) #f (car cachefiles))) ;; (and cachedir (conc cachedir "/" ".megatest.cfg-" megatest-version "-" megatest-fossil-hash))) (rccachef (if (null? cachefiles) #f (cdr cachefiles)))) ;; (and cachedir (conc cachedir "/" ".runconfigs.cfg-" megatest-version "-" megatest-fossil-hash))) ;; (cancreate (and cachedir (common:file-exists? cachedir)(file-write-access? cachedir) (not (common:in-running-test?))))) (set! *toppath* toppath) ;; This is needed when we are running as a test using CMDINFO as a datasource ;;(BB> "launch:setup-body -- cachefiles="cachefiles) (cond ;; if mtcachef exists just read it, however we need to assume toppath is available in $MT_RUN_AREA_HOME ((and (not force-reread) mtcachef rccachef use-cache (get-environment-variable "MT_RUN_AREA_HOME") (common:file-exists? mtcachef) (common:file-exists? rccachef)) ;;(BB> "launch:setup-body -- cond branch 1 - use-cache") (set! *configdat* (configf:read-alist mtcachef)) ;;(BB> "launch:setup-body -- 1 set! *configdat*="*configdat*) (set! *runconfigdat* (configf:read-alist rccachef)) (set! *configinfo* (list *configdat* (get-environment-variable "MT_RUN_AREA_HOME"))) (set! *configstatus* 'fulldata) (set! *toppath* (get-environment-variable "MT_RUN_AREA_HOME")) *toppath*) ;; there are no existing cached configs, do full reads of the configs and cache them ;; we have all the info needed to fully process runconfigs and megatest.config ((and ;; (not force-reread) ;; force-reread is irrelevant in the AND, could however OR it? mtcachef rccachef) ;; BB- why are we doing this without asking if caching is desired? ;;(BB> "launch:setup-body -- cond branch 2") (let* ((first-pass (configf:find-and-read-config ;; NB// sets MT_RUN_AREA_HOME as side effect mtconfig environ-patt: "env-override" given-toppath: toppath pathenvvar: "MT_RUN_AREA_HOME")) (first-rundat (let ((toppath (if toppath toppath (car first-pass)))) (configf:read-config ;; (conc toppath "/runconfigs.config") ;; this should be converted to runconfig:read but it is non-trivial, leaving it for now. (conc (if (string? toppath) toppath (get-environment-variable "MT_RUN_AREA_HOME")) "/runconfigs.config") *runconfigdat* #t sections: sections)))) (set! *runconfigdat* first-rundat) (if first-pass ;; (begin ;;(BB> "launch:setup-body -- \"first-pass\"=first-pass") (set! *configdat* (car first-pass)) ;;(BB> "launch:setup-body -- 2 set! *configdat*="*configdat*) (set! *configinfo* first-pass) (set! *toppath* (or toppath (cadr first-pass))) ;; use the gathered data unless already have it (set! toppath *toppath*) (if (not *toppath*) (begin (debug:print-error 0 *default-log-port* "you are not in a megatest area!") (exit 1))) (setenv "MT_RUN_AREA_HOME" *toppath*) ;; the seed read is done, now read runconfigs, cache it then read megatest.config one more time and cache it (let* ((keys (rmt:get-keys)) (key-vals (keys:target->keyval keys target)) (linktree (common:get-linktree)) ;; (or (getenv "MT_LINKTREE")(if *configdat* (configf:lookup *configdat* "setup" "linktree") #f))) ; (if *configdat* ; (configf:lookup *configdat* "setup" "linktree") ; (conc *toppath* "/lt")))) (second-pass (configf:find-and-read-config mtconfig environ-patt: "env-override" given-toppath: toppath pathenvvar: "MT_RUN_AREA_HOME")) (runconfigdat (begin ;; this read of the runconfigs will see any adjustments made by re-reading megatest.config (for-each (lambda (kt) (setenv (car kt) (cadr kt))) key-vals) (configf:read-config (conc toppath "/runconfigs.config") *runconfigdat* #t ;; consider using runconfig:read some day ... sections: sections))) (cachefiles (launch:get-cache-file-paths areapath toppath target mtconfig)) (mtcachef (car cachefiles)) (rccachef (cdr cachefiles))) ;; trap exception due to stale NFS handle -- Error: (open-output-file) cannot open file - Stale NFS file handle: "/p/fdk/gwa/lefkowit/mtTesting/qa/primbeqa/links/p1222/11/PDK_r1.1.1/prim/clean/pcell_testgen/.runconfigs.cfg-1.6427-7d1e789cb3f62f9cde719a4865bb51b3c17ea853" - ticket 220546342 ;; TODO - consider 1) using simple-lock to bracket cache write ;; 2) cache in hash on server, since need to do rmt: anyway to lock. (if rccachef (common:fail-safe (lambda () (configf:write-alist runconfigdat rccachef)) (conc "Could not write cache file - "rccachef))) (if mtcachef (common:fail-safe (lambda () (configf:write-alist *configdat* mtcachef)) (conc "Could not write cache file - "mtcachef))) (set! *runconfigdat* runconfigdat) (if (and rccachef mtcachef) (set! *configstatus* 'fulldata)))) ;; no configs found? should not happen but let's try to recover gracefully, return an empty hash-table (set! *configdat* (make-hash-table)) ))) ;; else read what you can and set the flag accordingly ;; here we don't have either mtconfig or rccachef (else ;;(BB> "launch:setup-body -- cond branch 3 - else") (let* ((cfgdat (configf:find-and-read-config (or (args:get-arg "-config") "megatest.config") environ-patt: "env-override" given-toppath: (get-environment-variable "MT_RUN_AREA_HOME") pathenvvar: "MT_RUN_AREA_HOME"))) (if (and cfgdat (list? cfgdat) (> (length cfgdat) 0) (hash-table? (car cfgdat))) (let* ((toppath (or (get-environment-variable "MT_RUN_AREA_HOME")(cadr cfgdat))) (rdat (configf:read-config (conc toppath ;; convert this to use runconfig:read! "/runconfigs.config") *runconfigdat* #t sections: sections))) (set! *configinfo* cfgdat) (set! *configdat* (car cfgdat)) (set! *runconfigdat* rdat) (set! *toppath* toppath) (set! *configstatus* 'partial)) (begin (debug:print-error 0 *default-log-port* "No " mtconfig " file found. Giving up.") (exit 2)))))) ;; COND ends here. ;; additional house keeping (let* ((linktree (or (common:get-linktree) (conc *toppath* "/lt")))) (if linktree (begin (if (not (common:file-exists? linktree)) (begin (handle-exceptions exn (begin (debug:print-error 0 *default-log-port* "Something went wrong when trying to create linktree dir at " linktree) (debug:print 0 *default-log-port* " message: " ((condition-property-accessor 'exn 'message) exn)) (exit 1)) (create-directory linktree #t)))) (handle-exceptions exn (begin (debug:print-error 0 *default-log-port* "Something went wrong when trying to create link to linktree at " *toppath*) (debug:print 0 *default-log-port* " message: " ((condition-property-accessor 'exn 'message) exn))) (let ((tlink (conc *toppath* "/lt"))) (if (not (common:file-exists? tlink)) (create-symbolic-link linktree tlink))))) (begin (debug:print-error 0 *default-log-port* "linktree not defined in [setup] section of megatest.config") ))) (if (and *toppath* (directory-exists? *toppath*)) (begin (setenv "MT_RUN_AREA_HOME" *toppath*) (setenv "MT_TESTSUITENAME" (common:get-testsuite-name))) (begin (debug:print-error 0 *default-log-port* "failed to find the top path to your Megatest area.") (set! *toppath* #f) ;; force it to be false so we return #f #f)) ;; one more attempt to cache the configs for future reading (let* ((cachefiles (launch:get-cache-file-paths areapath toppath target mtconfig)) (mtcachef (car cachefiles)) (rccachef (cdr cachefiles))) ;; trap exception due to stale NFS handle -- Error: (open-output-file) cannot open file - Stale NFS file handle: "...somepath.../.runconfigs.cfg-1.6427-7d1e789cb3f62f9cde719a4865bb51b3c17ea853" - ticket 220546342 ;; TODO - consider 1) using simple-lock to bracket cache write ;; 2) cache in hash on server, since need to do rmt: anyway to lock. (if (and rccachef *runconfigdat* (not (common:file-exists? rccachef))) (common:fail-safe (lambda () (configf:write-alist *runconfigdat* rccachef)) (conc "Could not write cache file - "rccachef)) ) (if (and mtcachef *configdat* (not (common:file-exists? mtcachef))) (common:fail-safe (lambda () (configf:write-alist *configdat* mtcachef)) (conc "Could not write cache file - "mtcachef)) ) (if (and rccachef mtcachef *runconfigdat* *configdat*) (set! *configstatus* 'fulldata))) ;; if have -append-config then read and append here (let ((cfname (args:get-arg "-append-config"))) (if (and cfname (file-read-access? cfname)) (configf:read-config cfname *configdat* #t))) ;; values are added to the hash, no need to do anything special. *toppath*))) (define (get-best-disk confdat testconfig) (let* ((disks (or (and testconfig (hash-table-ref/default testconfig "disks" #f)) (hash-table-ref/default confdat "disks" #f))) (minspace (let ((m (configf:lookup confdat "setup" "minspace"))) (string->number (or m "10000"))))) (if disks (let ((res (common:get-disk-with-most-free-space disks minspace))) ;; min size of 1000, seems tad dumb (if res (cdr res) (begin ;; (if (common:low-noise-print 20 "No valid disks or no disk with enough space") ;; (debug:print-error 0 *default-log-port* "No valid disks found in megatest.config. Please add some to your [disks] section and ensure the directory exists and has enough space!\n You can change minspace in the [setup] section of megatest.config. Current setting is: " minspace)) ;;(exit 1) (if (null? disks) (cons 1 (conc *toppath* "/runs")) (let ((paths (sort disks (lambda (x y) (> (string-length (cadr x)) (string-length (cadr y))))))) (let loop ((head (car paths)) (tail (cdr paths))) (let ((result (handle-exceptions exn #f (create-directory (cadr head) #t)))) (if result result (if (null? tail) (cons 1 (conc *toppath* "/runs")) (loop (car tail) (cdr tail)))))))))))))) ;; the code creates the necessary directories if it does not exist and returns the path. (define (launch:test-copy test-src-path test-path) (let* ((ovrcmd (let ((cmd (configf:lookup *configdat* "setup" "testcopycmd"))) (if cmd ;; substitute the TEST_SRC_PATH and TEST_TARG_PATH (string-substitute "TEST_TARG_PATH" test-path (string-substitute "TEST_SRC_PATH" test-src-path cmd #t) #t) #f))) (cmd (if ovrcmd ovrcmd (conc "rsync -av" (if (debug:debug-mode 1) "" "q") " " test-src-path "/ " test-path "/" " >> " test-path "/mt_launch.log 2>> " test-path "/mt_launch.log"))) (status (system cmd))) (if (not (eq? status 0)) (debug:print 2 *default-log-port* "ERROR: problem with running \"" cmd "\"")))) ;; Desired directory structure: ;; ;; <linkdir> - <target> - <testname> -. ;; | ;; v ;; <rundir> - <target> - <testname> -|- <itempath(s)> ;; ;; dir stored in test is: ;; ;; <linkdir> - <target> - <testname> [ - <itempath> ] ;; ;; All log file links should be stored relative to the top of link path ;; ;; <target> - <testname> [ - <itempath> ] ;; (define (create-work-area run-id run-info keyvals test-id test-src-path disk-path testname itemdat #!key (remtries 2)) (let* ((item-path (if (string? itemdat) itemdat (item-list->path itemdat))) ;; if pass in string - just use it (runname (if (string? run-info) ;; if we pass in a string as run-info use it as run-name. run-info (db:get-value-by-header (db:get-rows run-info) (db:get-header run-info) "runname"))) (contour #f) ;; NOT READY FOR THIS (args:get-arg "-contour")) ;; convert back to db: from rdb: - this is always run at server end (target (string-intersperse (map cadr keyvals) "/")) (not-iterated (equal? "" item-path)) ;; all tests are found at <rundir>/test-base or <linkdir>/test-base (testtop-base (conc target "/" runname "/" testname)) (test-base (conc testtop-base (if not-iterated "" "/") item-path)) ;; nb// if itempath is not "" then it is prefixed with "/" (toptest-path (conc disk-path (if contour (conc "/" contour) "") "/" testtop-base)) (test-path (conc disk-path (if contour (conc "/" contour) "") "/" test-base)) ;; ensure this exists first as links to subtests must be created there (linktree (common:get-linktree)) ;; WAS: (let ((rd (configf:lookup *configdat* "setup" "linktree"))) ;; (if rd rd (conc *toppath* "/runs")))) ;; which seems wrong ... (lnkbase (conc linktree (if contour (conc "/" contour) "") "/" target "/" runname)) (lnkpath (conc lnkbase "/" testname)) (lnkpathf (conc lnkpath (if not-iterated "" "/") item-path)) (lnktarget (conc lnkpath "/" item-path))) ;; Update the rundir path in the test record for all, rundir=physical, shortdir=logical ;; rundir shortdir (rmt:general-call 'test-set-rundir-shortdir run-id lnkpathf test-path testname item-path run-id) (debug:print 2 *default-log-port* "INFO:\n lnkbase=" lnkbase "\n lnkpath=" lnkpath "\n toptest-path=" toptest-path "\n test-path=" test-path) (if (not (common:file-exists? linktree)) (begin (debug:print 0 *default-log-port* "WARNING: linktree did not exist! Creating it now at " linktree) (create-directory linktree #t))) ;; (system (conc "mkdir -p " linktree)))) ;; create the directory for the tests dir links, this is needed no matter what... try up to three times (let loop ((done 3)) (let ((success (if (and (not (common:directory-exists? lnkbase)) (not (common:file-exists? lnkbase))) (handle-exceptions exn (begin (debug:print-error 0 *default-log-port* "Problem creating linktree base at " lnkbase) (print-error-message exn (current-error-port)) #t) (create-directory lnkbase #t) #f)))) (if (and (not success)(> done 0)) (loop (- done 1))))) ;; update the toptest record with its location rundir, cache the path ;; This wass highly inefficient, one db write for every subtest, potentially ;; thousands of unnecessary updates, cache the fact it was set and don't set it ;; again. ;; Now create the link from the test path to the link tree, however ;; if the test is iterated it is necessary to create the parent path ;; to the iteration. use pathname-directory to trim the path by one ;; level (if (not not-iterated) ;; i.e. iterated (let ((iterated-parent (pathname-directory (conc lnkpath "/" item-path)))) (debug:print-info 2 *default-log-port* "Creating iterated parent " iterated-parent) (handle-exceptions exn (begin (debug:print-error 0 *default-log-port* " Failed to create directory " iterated-parent ((condition-property-accessor 'exn 'message) exn) ", exiting") (exit 1)) (create-directory iterated-parent #t)))) (if (symbolic-link? lnkpath) (handle-exceptions exn (begin (debug:print-error 0 *default-log-port* " Failed to remove symlink " lnkpath ((condition-property-accessor 'exn 'message) exn) ", exiting") (exit 1)) (delete-file lnkpath))) (if (not (or (common:file-exists? lnkpath) (symbolic-link? lnkpath))) (handle-exceptions exn (begin (debug:print-error 0 *default-log-port* " Failed to create symlink " lnkpath ((condition-property-accessor 'exn 'message) exn) ", exiting") (exit 1)) (create-symbolic-link toptest-path lnkpath))) ;; NB - This was not working right - some top tests are not getting the path set!!! ;; ;; Do the setting of this record after the paths are created so that the shortdir can ;; be set to the real directory location. This is safer for future clean up if the link ;; tree is damaged or lost. ;; (if (not (hash-table-ref/default *toptest-paths* testname #f)) (let* ((testinfo (rmt:get-test-info-by-id run-id test-id)) ;; run-id testname item-path)) (curr-test-path (if testinfo ;; (filedb:get-path *fdb* ;; (db:get-path dbstruct ;; (rmt:sdb-qry 'getstr (db:test-get-rundir testinfo) ;; ) ;; ) #f))) (hash-table-set! *toptest-paths* testname curr-test-path) ;; NB// Was this for the test or for the parent in an iterated test? (rmt:general-call 'test-set-rundir-shortdir run-id lnkpath (if (common:file-exists? lnkpath) ;; (resolve-pathname lnkpath) (common:nice-path lnkpath) lnkpath) testname "" run-id) ;; (rmt:general-call 'test-set-rundir run-id lnkpath testname "") ;; toptest-path) (if (or (not curr-test-path) (not (directory-exists? toptest-path))) (begin (debug:print-info 2 *default-log-port* "Creating " toptest-path " and link " lnkpath) (handle-exceptions exn #f ;; don't care to catch and deal with errors here for now. (create-directory toptest-path #t)) (hash-table-set! *toptest-paths* testname toptest-path))))) ;; The toptest path has been created, the link to the test in the linktree has ;; been created. Now, if this is an iterated test the real test dir must be created (if (not not-iterated) ;; this is an iterated test (begin ;; (let ((lnktarget (conc lnkpath "/" item-path))) (debug:print 2 *default-log-port* "Setting up sub test run area") (debug:print 2 *default-log-port* " - creating run area in " test-path) (handle-exceptions exn (begin (debug:print-error 0 *default-log-port* " Failed to create directory " test-path ((condition-property-accessor 'exn 'message) exn) ", exiting") (exit 1)) (create-directory test-path #t)) (debug:print 2 *default-log-port* " - creating link from: " test-path "\n" " to: " lnktarget) ;; If there is already a symlink delete it and recreate it. (handle-exceptions exn (begin (debug:print-error 0 *default-log-port* " Failed to re-create link " lnktarget ((condition-property-accessor 'exn 'message) exn) ", exiting") (exit)) (if (symbolic-link? lnktarget) (delete-file lnktarget)) (if (not (common:file-exists? lnktarget)) (create-symbolic-link test-path lnktarget))))) (if (not (directory? test-path)) (create-directory test-path #t)) ;; this is a hack, I don't know why out of the blue this path does not exist sometimes (if (and test-src-path (directory? test-path)) (begin (launch:test-copy test-src-path test-path) (list lnkpathf lnkpath )) (if (and test-src-path (> remtries 0)) (begin (debug:print-error 0 *default-log-port* "Failed to create work area at " test-path " with link at " lnktarget ", remaining attempts " remtries) ;; (create-work-area run-id run-info keyvals test-id test-src-path disk-path testname itemdat remtries: (- remtries 1))) (list #f #f))))) (define (launch:handle-zombie-tests run-id) (let* ((key (conc "zombiescan-runid-"run-id)) (now (current-seconds)) (threshold (- (current-seconds) (* 2 (or (configf:lookup-number *configdat* "setup" "deadtime") 120)))) (val (rmt:get-var key)) (do-scan? (cond ((not val) #t) ((< val threshold) #t) (else #f)))) (when do-scan? (debug:print 1 *default-log-port* "INFO: search and mark zombie tests") (rmt:set-var key (current-seconds)) (rmt:find-and-mark-incomplete run-id #f)))) ;; 1. look though disks list for disk with most space ;; 2. create run dir on disk, path name is meaningful ;; 3. create link from run dir to megatest runs area ;; 4. remotely run the test on allocated host ;; - could be ssh to host from hosts table (update regularly with load) ;; - could be netbatch ;; (launch-test db (cadr status) test-conf)) (define (launch-test test-id run-id run-info keyvals runname test-conf test-name test-path itemdat params) (mutex-lock! *launch-setup-mutex*) ;; setting variables and processing the testconfig is NOT thread-safe, reuse the launch-setup mutex (let* ( ;; (lock-key (conc "test-" test-id)) ;; (got-lock (let loop ((lock (rmt:no-sync-get-lock lock-key)) ;; (expire-time (+ (current-seconds) 15))) ;; give up on getting the lock and steal it after 15 seconds ;; (if (car lock) ;; #t ;; (if (> (current-seconds) expire-time) ;; (begin ;; (debug:print-info 0 *default-log-port* "Timed out waiting for a lock to launch test " keyvals " " runname " " test-name " " test-path) ;; (rmt:no-sync-del! lock-key) ;; destroy the lock ;; (loop (rmt:no-sync-get-lock lock-key) expire-time)) ;; ;; (begin ;; (thread-sleep! 1) ;; (loop (rmt:no-sync-get-lock lock-key) expire-time)))))) (item-path (item-list->path itemdat)) (contour #f)) ;; NOT READY FOR THIS (args:get-arg "-contour"))) (let loop ((delta (- (current-seconds) *last-launch*)) (launch-delay (configf:lookup-number *configdat* "setup" "launch-delay" default: 1))) (if (> launch-delay delta) (begin (if (common:low-noise-print 1200 "test launch delay") ;; every two hours or so remind the user about launch delay. (debug:print-info 0 *default-log-port* "NOTE: test launches are delayed by " launch-delay " seconds. See megatest.config launch-delay setting to adjust.")) ;; launch of " test-name " for " (- launch-delay delta) " seconds")) (thread-sleep! (- launch-delay delta)) (loop (- (current-seconds) *last-launch*) launch-delay)))) (change-directory *toppath*) (alist->env-vars ;; consolidate this code with the code in megatest.scm for "-execute", *maybe* - the longer they are set the longer each launch takes (must be non-overlapping with the vars) (append (list (list "MT_RUN_AREA_HOME" *toppath*) (list "MT_TEST_NAME" test-name) (list "MT_RUNNAME" runname) (list "MT_ITEMPATH" item-path) (list "MT_CONTOUR" contour) ) itemdat)) (let* ((tregistry (tests:get-all)) ;; third param (below) is system-allowed ;; for tconfig, why do we allow fallback to test-conf? (tconfig (or (tests:get-testconfig test-name item-path tregistry #t force-create: #t) (begin (debug:print 0 *default-log-port* "WARNING: falling back to pre-calculated testconfig. This is likely not desired.") test-conf))) ;; force re-read now that all vars are set (useshell (let ((ush (configf:lookup *configdat* "jobtools" "useshell"))) (if ush (if (equal? ush "no") ;; must use "no" to NOT use shell #f ush) #t))) ;; default is yes (runscript (configf:lookup tconfig "setup" "runscript")) (ezsteps (> (length (hash-table-ref/default tconfig "ezsteps" '())) 0)) ;; don't send all the steps, could be big, just send a flag (subrun (> (length (hash-table-ref/default tconfig "subrun" '())) 0)) ;; send a flag to process a subrun ;; (diskspace (configf:lookup tconfig "requirements" "diskspace")) ;; (memory (configf:lookup tconfig "requirements" "memory")) ;; (hosts (configf:lookup *configdat* "jobtools" "workhosts")) ;; I'm pretty sure this was never completed (remote-megatest (configf:lookup *configdat* "setup" "executable")) (run-time-limit (or (configf:lookup tconfig "requirements" "runtimelim") (configf:lookup *configdat* "setup" "runtimelim"))) ;; FIXME SOMEDAY: not good how this is so obtuse, this hack is to ;; allow running from dashboard. Extract the path ;; from the called megatest and convert dashboard ;; or dboard to megatest (local-megatest (let* ((lm (car (argv))) (dir (pathname-directory lm)) (exe (pathname-strip-directory lm))) (conc (if dir (conc dir "/") "") (case (string->symbol exe) ((dboard) "../megatest") ((mtest) "../megatest") ((dashboard) "megatest") (else exe))))) (launcher (common:get-launcher *configdat* test-name item-path)) ;; (configf:lookup *configdat* "jobtools" "launcher")) (test-sig (conc (common:get-testsuite-name) ":" test-name ":" item-path)) ;; (item-list->path itemdat))) ;; test-path is the full path including the item-path (work-area #f) (toptest-work-area #f) ;; for iterated tests the top test contains data relevant for all (diskpath #f) (cmdparms #f) (fullcmd #f) ;; (define a (with-output-to-string (lambda ()(write x)))) (mt-bindir-path #f) (testinfo (rmt:get-test-info-by-id run-id test-id)) (mt_target (string-intersperse (map cadr keyvals) "/")) (debug-param (append (if (args:get-arg "-debug") (list "-debug" (args:get-arg "-debug")) '()) (if (args:get-arg "-logging")(list "-logging") '())))) ;; (if hosts (set! hosts (string-split hosts))) ;; set the megatest to be called on the remote host (if (not remote-megatest)(set! remote-megatest local-megatest)) ;; "megatest")) (set! mt-bindir-path (pathname-directory remote-megatest)) (if launcher (set! launcher (string-split launcher))) ;; set up the run work area for this test (if (and (args:get-arg "-preclean") ;; user has requested to preclean for this run (not (member (db:test-get-rundir testinfo)(list "n/a" "/tmp/badname")))) ;; n/a is a placeholder and thus not a read dir (begin (debug:print-info 0 *default-log-port* "attempting to preclean directory " (db:test-get-rundir testinfo) " for test " test-name "/" item-path) (runs:remove-test-directory testinfo 'remove-data-only))) ;; remove data only, do not perturb the record ;; prevent overlapping actions - set to LAUNCHED as early as possible ;; ;; the following call handles waiver propogation. cannot yet condense into roll-up-pass-fail (tests:test-set-status! run-id test-id "LAUNCHED" "n/a" #f #f) ;; (if launch-results launch-results "FAILED")) (rmt:set-state-status-and-roll-up-items run-id test-name item-path #f "LAUNCHED" #f) ;; (pp (hash-table->alist tconfig)) (set! diskpath (get-best-disk *configdat* tconfig)) (if diskpath (let ((dat (create-work-area run-id run-info keyvals test-id test-path diskpath test-name itemdat))) (set! work-area (car dat)) (set! toptest-work-area (cadr dat)) (debug:print-info 2 *default-log-port* "Using work area " work-area)) (begin (set! work-area (conc test-path "/tmp_run")) (create-directory work-area #t) (debug:print 0 *default-log-port* "WARNING: No disk work area specified - running in the test directory under tmp_run"))) (set! cmdparms (base64:base64-encode (z3:encode-buffer (with-output-to-string (lambda () ;; (list 'hosts hosts) (write (list (list 'testpath test-path) ;; (list 'transport (conc *transport-type*)) ;; (list 'serverinf *server-info*) (list 'homehost (let* ((hhdat (common:get-homehost))) (if hhdat (car hhdat) #f))) (list 'serverurl (if *runremote* (remote-server-url *runremote*) #f)) ;; (list 'areaname (common:get-testsuite-name)) (list 'toppath *toppath*) (list 'work-area work-area) (list 'test-name test-name) (list 'runscript runscript) (list 'run-id run-id ) (list 'test-id test-id ) ;; (list 'item-path item-path ) (list 'itemdat itemdat ) (list 'megatest remote-megatest) (list 'ezsteps ezsteps) (list 'subrun subrun) (list 'target mt_target) (list 'contour contour) (list 'runtlim (if run-time-limit (common:hms-string->seconds run-time-limit) #f)) (list 'env-ovrd (hash-table-ref/default *configdat* "env-override" '())) (list 'set-vars (if params (hash-table-ref/default params "-setvars" #f))) (list 'runname runname) (list 'mt-bindir-path mt-bindir-path)))))))) ;; clean out step records from previous run if they exist ;; (rmt:delete-test-step-records run-id test-id) ;; if the dir does not exist we may have a itempath where individual variables are a path, launch anyway (if (common:file-exists? work-area) (change-directory work-area)) ;; so that log files from the launch process don't clutter the test dir (cond ;; ((and launcher hosts) ;; must be using ssh hostname ;; (set! fullcmd (append launcher (car hosts)(list remote-megatest "-m" test-sig "-execute" cmdparms) debug-param))) ;; (set! fullcmd (append launcher (car hosts)(list remote-megatest test-sig "-execute" cmdparms)))) (launcher (set! fullcmd (append launcher (list remote-megatest "-m" test-sig "-execute" cmdparms) debug-param))) ;; (set! fullcmd (append launcher (list remote-megatest test-sig "-execute" cmdparms)))) (else (if (not useshell)(debug:print 0 *default-log-port* "WARNING: internal launching will not work well without \"useshell yes\" in your [jobtools] section")) (set! fullcmd (append (list remote-megatest "-m" test-sig "-execute" cmdparms) debug-param (list (if useshell "&" "")))))) ;; (set! fullcmd (list remote-megatest test-sig "-execute" cmdparms (if useshell "&" ""))))) (if (args:get-arg "-xterm")(set! fullcmd (append fullcmd (list "-xterm")))) (debug:print 1 *default-log-port* "Launching " work-area) ;; set pre-launch-env-vars before launching, keep the vars in prevvals and put the envionment back when done (debug:print 4 *default-log-port* "fullcmd: " fullcmd) (set! *last-launch* (current-seconds)) ;; all that junk above takes time, set this as late as possible. (let* ((commonprevvals (alist->env-vars (hash-table-ref/default *configdat* "env-override" '()))) (miscprevvals (alist->env-vars ;; consolidate this code with the code in megatest.scm for "-execute" (append (list (list "MT_TEST_RUN_DIR" work-area) (list "MT_TEST_NAME" test-name) (list "MT_ITEM_INFO" (conc itemdat)) (list "MT_RUNNAME" runname) (list "MT_TARGET" mt_target) (list "MT_ITEMPATH" item-path) ) itemdat))) (testprevvals (alist->env-vars (hash-table-ref/default tconfig "pre-launch-env-overrides" '()))) ;; Launchwait defaults to true, must override it to turn off wait (launchwait (if (equal? (configf:lookup *configdat* "setup" "launchwait") "no") #f #t)) (launch-results-prev (apply (if launchwait ;; BB: TODO: refactor this to examine return code of launcher, if nonzero, set state to launch failed. process:cmd-run-with-stderr-and-exitcode->list process-run) (if useshell (let ((cmdstr (string-intersperse fullcmd " "))) (if launchwait cmdstr (conc cmdstr " >> mt_launch.log 2>&1 &"))) (car fullcmd)) (if useshell '() (cdr fullcmd)))) (success (if launchwait (equal? 0 (cadr launch-results-prev)) #t)) (launch-results (if launchwait (car launch-results-prev) launch-results-prev))) (if (not success) (tests:test-set-status! run-id test-id "COMPLETED" "DEAD" "launcher failed; exited non-zero; check mt_launch.log" #f)) ;; (if launch-results launch-results "FAILED")) (mutex-unlock! *launch-setup-mutex*) ;; yes, really should mutex all the way to here. Need to put this entire process into a fork. ;; (rmt:no-sync-del! lock-key) ;; release the lock for starting this test (if (not launchwait) ;; give the OS a little time to allow the process to start (thread-sleep! 0.01)) (with-output-to-file "mt_launch.log" (lambda () (print "LAUNCHCMD: " (string-intersperse fullcmd " ")) (if (list? launch-results) (apply print launch-results) (print "NOTE: launched \"" fullcmd "\"\n but did not wait for it to proceed. Add the following to megatest.config \n[setup]\nlaunchwait yes\n if you have problems with this")) #:append)) (debug:print 2 *default-log-port* "Launching completed, updating db") (debug:print 2 *default-log-port* "Launch results: " launch-results) (if (not launch-results) (begin (print "ERROR: Failed to run " (string-intersperse fullcmd " ") ", exiting now") ;; (sqlite3:finalize! db) ;; good ole "exit" seems not to work ;; (_exit 9) ;; but this hack will work! Thanks go to Alan Post of the Chicken email list ;; NB// Is this still needed? Should be safe to go back to "exit" now? (process-signal (current-process-id) signal/kill) )) (alist->env-vars miscprevvals) (alist->env-vars testprevvals) (alist->env-vars commonprevvals) launch-results)) (change-directory *toppath*))) ;; recover a test where the top controlling mtest may have died ;; (define (launch:recover-test run-id test-id) ;; this function is called on the test run host via ssh ;; ;; 1. look at the process from pid ;; - is it owned by calling user ;; - it it's run directory correct for the test ;; - is there a controlling mtest (maybe stuck) ;; 2. if recovery is needed watch pid ;; - when it exits take the exit code and do the needful ;; (let* ((pid (rmt:test-get-top-process-pid run-id test-id)) (psres (with-input-from-pipe (conc "ps -F -u " (current-user-name) " | grep -E '" pid " ' | grep -v 'grep -E " pid "'") (lambda () (read-line)))) (rundir (if (string? psres) ;; real process owned by user (read-symbolic-link (conc "/proc/" pid "/cwd")) #f))) ;; now wait on that process if all is correct ;; periodically update the db with runtime ;; when the process exits look at the db, if still RUNNING after 10 seconds set ;; state/status appropriately (process-wait pid))) ;; Do not rpc this one, do the underlying calls!!! (define (tests:test-set-status! run-id test-id state status comment dat #!key (work-area #f)) (let* ((real-status status) (otherdat (if dat dat (make-hash-table))) (testdat (rmt:get-test-info-by-id run-id test-id)) (test-name (db:test-get-testname testdat)) (item-path (db:test-get-item-path testdat)) ;; before proceeding we must find out if the previous test (where all keys matched except runname) ;; was WAIVED if this test is FAIL ;; NOTES: ;; 1. Is the call to test:get-previous-run-record remotified? ;; 2. Add test for testconfig waiver propagation control here ;; (prev-test (if (equal? status "FAIL") (rmt:get-previous-test-run-record run-id test-name item-path) #f)) (waived (if prev-test (if prev-test ;; true if we found a previous test in this run series (let ((prev-status (db:test-get-status prev-test)) (prev-state (db:test-get-state prev-test)) (prev-comment (db:test-get-comment prev-test))) (debug:print 4 *default-log-port* "prev-status " prev-status ", prev-state " prev-state ", prev-comment " prev-comment) (if (and (equal? prev-state "COMPLETED") (equal? prev-status "WAIVED")) (if comment comment prev-comment) ;; waived is either the comment or #f #f)) #f) #f))) (if (and waived (tests:check-waiver-eligibility testdat prev-test)) (set! real-status "WAIVED")) (debug:print 4 *default-log-port* "real-status " real-status ", waived " waived ", status " status) ;; update the primary record IF state AND status are defined (if (and state status) (begin (rmt:set-state-status-and-roll-up-items run-id test-id item-path state real-status (if waived waived comment)) ;; (mt:process-triggers run-id test-id state real-status) ;; triggers are called in test-set-state-status )) ;; if status is "AUTO" then call rollup (note, this one modifies data in test ;; run area, it does remote calls under the hood. ;; (if (and test-id state status (equal? status "AUTO")) ;; (rmt:test-data-rollup run-id test-id status)) ;; add metadata (need to do this way to avoid SQL injection issues) ;; :first_err ;; (let ((val (hash-table-ref/default otherdat ":first_err" #f))) ;; (if val ;; (sqlite3:execute db "UPDATE tests SET first_err=? WHERE run_id=? AND testname=? AND item_path=?;" val run-id test-name item-path))) ;; ;; ;; :first_warn ;; (let ((val (hash-table-ref/default otherdat ":first_warn" #f))) ;; (if val ;; (sqlite3:execute db "UPDATE tests SET first_warn=? WHERE run_id=? AND testname=? AND item_path=?;" val run-id test-name item-path))) (let ((category (hash-table-ref/default otherdat ":category" "")) (variable (hash-table-ref/default otherdat ":variable" "")) (value (hash-table-ref/default otherdat ":value" #f)) (expected (hash-table-ref/default otherdat ":expected" "n/a")) (tol (hash-table-ref/default otherdat ":tol" "n/a")) (units (hash-table-ref/default otherdat ":units" "")) (type (hash-table-ref/default otherdat ":type" "")) (dcomment (hash-table-ref/default otherdat ":comment" ""))) (debug:print 4 *default-log-port* "category: " category ", variable: " variable ", value: " value ", expected: " expected ", tol: " tol ", units: " units) (if (and value) ;; require only value; BB was- all three required (let ((dat (conc category "," variable "," value "," expected "," tol "," units "," dcomment ",," ;; extra comma for status type ))) ;; This was run remote, don't think that makes sense. Perhaps not, but that is the easiest path for the moment. (rmt:csv->test-data run-id test-id dat) ;; This was added in check-in a5adfa3f9a. Message was: "...added delay in set-values to allow for delayed write on server start" ;; I'm inserting an arbitrary rmt: call to force/ensure that the server is available to (hopefully) prevent a communication issue. (rmt:get-var "MEGATEST_VERSION") ;; this does NOTHING but ensure the server is reachable. This is almost certainly NOT needed :) ;; BB - commentiong out arbitrary 10 second wait (thread-sleep! 10) ;; add 10 second delay before quit incase rmt needs time to start a server. ))) ;; need to update the top test record if PASS or FAIL and this is a subtest ;;;;;; (if (not (equal? item-path "")) ;;;;;; (rmt:set-state-status-and-roll-up-items run-id test-name item-path state status #f) ;;;;;) (if (or (and (string? comment) (string-match (regexp "\\S+") comment)) waived) (let ((cmt (if waived waived comment))) (rmt:general-call 'set-test-comment run-id cmt test-id))))) ) |
Modified runsmod.scm from [bd0c648c0b] to [a46fc03039].
︙ | |||
138 139 140 141 142 143 144 | 138 139 140 141 142 143 144 145 146 147 148 149 150 151 | - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - | (configf:section-vars *configdat* "env-override")))) |
︙ | |||
2688 2689 2690 2691 2692 2693 2694 2695 | 2601 2602 2603 2604 2605 2606 2607 2608 2609 2610 2611 2612 2613 2614 2615 2616 2617 2618 2619 2620 2621 2622 2623 2624 2625 2626 2627 2628 2629 2630 2631 2632 2633 2634 2635 2636 2637 2638 2639 2640 2641 2642 2643 2644 2645 2646 2647 2648 2649 2650 2651 2652 2653 2654 | + + + + + + + + + + + + + + + + + + + + + - + + + + + + + + + + + + + + + + + + + + + + + + + | (delete-file f))) files)))) (debug:print-error 0 *default-log-port* "-clean-cache requires -runname.")) (debug:print-error 0 *default-log-port* "-clean-cache requires -target or -reqtarg"))) ;; Spec for End of test ;; At end of each test call, after marking self as COMPLETED do run-state-status-rollup ;; At transition to run COMPLETED/X do hooks ;; Definition: test_dead if event_time + duration + 1 minute? < current_time AND ;; we can prove the process is not alive (ssh host pstree -A pid) ;; if dead safe to mark the test as killed in the db ;; State/status table ;; new ;; 100% COMPLETED/ (PASS,FAIL,ABORT etc.) ==> COMPLETED / X where X is same as itemized rollup ;; > 3 RUNNING with not test_dead do nothing (run should already be RUNNING/ na ;; > 0 RUNNING and test_dead then send KILLREQ ==> COMPLETED ;; 0 RUNNING ==> this is actually the first condition, should not get here (define (launch:end-of-run-check run-id ) (let* ((not-completed-cnt (rmt:get-not-completed-cnt run-id)) (running-cnt (rmt:get-count-tests-running-for-run-id run-id)) (all-test-launched (rmt:get-var (conc "lunch-complete-" run-id))) (current-state (rmt:get-run-state run-id)) (current-status (rmt:get-run-status run-id))) ;;get-vars run-id to query metadata table to check if all completed. if all-test-launched = yes then only not-completed-cnt = 0 means everyting is completed if no entry found in the table do nothing (debug:print 0 *default-log-port* "rollup run state/status") (rmt:set-state-status-and-roll-up-run run-id current-state current-status) |
Modified subrunmod.scm from [8f236dbe89] to [cc90746185].
︙ | |||
23 24 25 26 27 28 29 | 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 | - + | (declare (uses mtconfigf)) (module subrunmod * (import scheme chicken data-structures extras) (import (prefix sqlite3 sqlite3:) posix typed-records srfi-18 |
︙ |