50 descendants of a097994a72e93993
|
2017-04-04
| ||
| 17:38 | added '-prepend-contour', which allows for optional prepending of MT_CONTOUR/MT_AREA/ to target check-in: d9f734d9b8 user: srehman tags: v1.64-prepend-contour, v1.64-envdebug | |
| 14:54 | Hacks to make newdashboard start - just a jumpstarter check-in: 805f902745 user: mrwellan tags: v1.64-envdebug | |
| 12:08 | Added Roberts nbfake changes, use a dotfile for the db ready file flag check-in: b9bcceaeae user: mrwellan tags: v1.64-envdebug, v1.6403 | |
|
2017-04-03
| ||
| 22:05 | Joined transactions for initiating databases. This problably fixes many issues! check-in: b158f5ed64 user: matt tags: v1.64-envdebug | |
|
2017-03-31
| ||
| 16:05 | Added config option dashboard/poll-interval (in milliseconds) for dashboard updates check-in: e7a6790b9b user: mrwellan tags: v1.64-envdebug | |
|
2017-03-30
| ||
| 12:30 | v1.64-cache-issue check-in: e2ffc3801e user: bjbarcla tags: v1.64-envdebug | |
|
2017-03-29
| ||
| 11:03 | added unlock_db.sh to utils check-in: fd47a3e816 user: bjbarcla tags: v1.64 | |
| 10:50 | added utilities I developed on the side to assist in transition. Not functional yet in this area. check-in: 275dbf2c2e user: bjbarcla tags: v1.64 | |
|
2017-03-28
| ||
| 18:46 | Added signature for FATAL ERROR: to triage.rb check-in: 4e774e65ea user: matt tags: v1.64 | |
| 18:35 | special handler for strange configdat not defined issue. Updated to retry up to five times. check-in: 9b63886415 user: matt tags: v1.64 | |
| 17:07 | adding triage script check-in: 01756c020e user: bjbarcla tags: v1.64 | |
| 14:59 | Fixed several unusual crashes check-in: 51fbce80b9 user: matt tags: v1.64 | |
| 13:06 | Fixed call to runs:clear-cache with wrong arguments. Make tests:get-tests-search-path resistant to bad config data check-in: 619330088c user: matt tags: v1.64 | |
| 11:36 | Add small delay on first call to wait-on-server before trying to start a server. check-in: 723b985766 user: matt tags: v1.64 | |
| 10:54 | Force starting a server and wait for it when launching runs. This prevents server run-away but doesn't fix the underlying issue. check-in: 27b1636e7b user: matt tags: v1.64 | |
| 10:48 | Force starting a server and wait for it when launching runs. This prevents server run-away but doesn't fix the underlying issue. check-in: ba2401c3f6 user: matt tags: v1.64 | |
| 08:39 | Fixed unit test running. To try do: cd tests;make all-rmt.log. Improved information from a server-side crash. check-in: 263cdeb6eb user: matt tags: v1.64 | |
| 01:50 | Bits of refactoring to simplify rmt.scm check-in: 2158d4a3b0 user: matt tags: v1.64 | |
| 01:01 | should be resetting the connection struct when connection is failed to make check-in: 333f542a2e user: matt tags: v1.64 | |
| 00:35 | Improved reliability but now have issue with connection. check-in: 45da129709 user: matt tags: v1.64 | |
|
2017-03-27
| ||
| 23:59 | protected config file call to delete-file with exception handler. Fixed logic on connecting using CMDINFO. Fixed -list-servers and -kill-servers. Turned exception handler back on in portlogger. Removed the addition of a little noise from the server timeout handling in rmt.scm check-in: 7b318f91bd user: matt tags: v1.64 | |
| 12:41 | prereq proc is supposed to return test records, not test names. check-in: 2b9e485b7f user: matt tags: v1.64 | |
| 12:05 | Improvements to exclusive mode (but it still isn't working quite right). Factored clean-cache into a reusable function and called it also in -rerun*. Reduced number of server files to analyze. This may reduce the probability of a runaway server situation. check-in: 0f50a21b44 user: matt tags: v1.64 | |
| 00:59 | Clone always for fossil sensing. Need additional logic otherwise check-in: dc0c36c096 user: matt tags: v1.64 | |
| 00:42 | First attempt at the exclusive mode. Not working yet but leaving the code on the v1.64 branch in anticipation of fixing it. check-in: 3269b1925d user: matt tags: v1.64 | |
|
2017-03-26
| ||
| 21:58 | Added -preclean to all automatic run calls. Might need to be able to switch that off in future. runtrans=auto uses automatically generated run name, runtrans with spec in *runname-mappers* will use it, otherwise the runtrans is used as the actual runnanme. This is needed for run-in-place reruns check-in: 6b8e58d2d1 user: matt tags: v1.64 | |
| 17:47 | Added browsing of logs to RH Button for tests from runs view buttons. Added (speculative) fix for unstable sort on steps. check-in: 89dd44cd11 user: matt tags: v1.64 | |
|
2017-03-24
| ||
| 23:35 | Merged in multi-server-hack check-in: 99a40be38a user: matt tags: v1.64 | |
| 22:39 | Merged v1.63 changes into v1.64 check-in: b2e6452e2f user: matt tags: v1.64 | |
| 22:18 | non-good commit of merge multi-server-hack into v1.64 Leaf check-in: 42e2fcd1c8 user: matt tags: private (unpublished) | |
| 18:22 | Fixed couple issues with the throttle by api load but still not sure it is working right. Cleaned up few other transport things. Closed-Leaf check-in: e7dcebe686 user: matt tags: multi-server-hack | |
| 15:55 | Default to treating a log as recent instead of ancient. This might be part of the run-away servers issue check-in: 31427e83dd user: matt tags: multi-server-hack | |
| 15:39 | merged in protection against too demanding clients Leaf check-in: 57e72d2e76 user: bjbarcla tags: v1.63 | |
| 15:18 | Reduced threshold for server pushback on clients to 20 api calls/parallel. Added couple more calls to check if server is running before starting a new one. check-in: 37e149d637 user: matt tags: multi-server-hack | |
| 14:52 | Tell calling client to wait and try again if server is overloaded (in terms of parallel api calls over 25) check-in: fd3c06195d user: matt tags: multi-server-hack | |
| 13:51 | Show connection stats every 60 seconds. Remove stat of megatest.db from rmt:send-receive, it was happening on every call. check-in: 6baac6187e user: matt tags: multi-server-hack | |
| 11:27 | Merged v1.63 changes to multi-server-hack check-in: 8a6ca9fd18 user: matt tags: multi-server-hack | |
| 11:19 | Added some error handling on the locking calls. Silenced a dashboard message. check-in: a5dbcdd2ac user: matt tags: v1.63 | |
|
2017-03-23
| ||
| 17:50 | Couple fixes for variable server hack check-in: 842f12e5fe user: matt tags: multi-server-hack | |
| 16:52 | Hack for variable number of servers (default 3) check-in: e86b57ccb0 user: matt tags: multi-server-hack | |
| 13:56 | Added message when server is forced Closed-Leaf check-in: 6a476e9ca7 user: matt tags: runaway-servers-fix | |
| 13:38 | merged lockfile fix check-in: f25594cb64 user: bjbarcla tags: v1.63 | |
| 13:27 | Remove -O4, fixed the force server switch. check-in: c5f5a4ad19 user: matt tags: runaway-servers-fix | |
| 12:49 | adding some file-exists? protections and a fix to force-server? from 1.63 check-in: 211ecbabeb user: bjbarcla tags: runaway-servers-fix | |
| 12:23 | merged in matts fix for testsuite; resolved conflicts with common:file-exists? additions check-in: 474e1543fa user: bjbarcla tags: v1.63 | |
| 12:12 | fixed let in common:force-server? check-in: 4df9fd5d18 user: bjbarcla tags: runaway-servers-fix | |
| 11:34 | Fixed wrong testsuite name calculation problem. check-in: 8c653071c7 user: matt tags: runaway-servers-fix | |
| 01:25 | fixed bug in common:force-server? check-in: 7a06fb3f65 user: bjbarcla tags: v1.63 | |
| 01:18 | merged runaway servers fix check-in: c80ac44cb9 user: bjbarcla tags: v1.63 | |
| 00:53 | Check that server isn't already running before doing kind-run. This was working fine with low load but when load got high it would run away. check-in: a097994a72 user: matt tags: runaway-servers-fix | |