2018.1 build priority rules.
Does any one know what the current rules for how build priority is calculated in TeamCity 2018.1.*? I have multiple long regression test build configurations that I would like to not run for a while with there are other shorter jobs waiting. These regression build configurations take 6-8 hour each and have a build priority of -100 by the rules given in Build-priorities-mechanism these settings should cause these long builds to give what for several weeks to the other zero priority builds, yet the long builds seem to start having positive priority after less than one day.
The post was last updated in the TC 8.* error so maybe the rules have changed I am using TC 2018.1.2?
I am current trying to set teamcity.buildqueue.priorityWeight and teamcity.buildqueue.waitWeight to see if this changes the behavior.
Is there a more recent reference?
Thank to JetBrains for there great product.
Please sign in to leave a comment.
Hi Chris,
while behavior might have been slightly adjusted here or there, there shouldn't have been significant changes. As far as I can tell, the internal properties should still work. Did you observe a change in behavior since adding them?
Hi Denis,
Thanks for the information.
I have not yet observed any changes, but since your help on reducing the number of runs of the long build processes we don't have the multiple days of work in the build queue that often, may be once or twice a month, so I have not yet had the conditions to observe. At the moment I am using the build priority rules to have a graceful degradation in performs during high need periods when the primary goal is to have all related builds for a VCS check-in done in about 24 hours cannot be met.
I will report back when I get a good observation.
Setting
Hi Chris,
it's important to understand that the priorities and weights are... just that. TeamCity should use them internally for calculating build order, but there are other factors taken into account to avoid starvation of builds. Keep us posted with the results of your tests.