Azure not being run after feedstock creation

For the last two feedstocks I created via staged-recipes, only Appveyor was run on the feedstock after its creation. After noticing this, I added a single space to the meta.yaml to trigger the CI again and then the normal Azure processes were triggered and the packages actually built. The two feedstocks that I know were affected by this are:

• r-seqminer
• r-fateid

[jakirkham]

Hi Leah 👋 Thanks for coming by.

Recently we saw this with commit ( https://github.com/conda-forge/ogre-feedstock/commit/e96f50b94a633309f610eafa1f189dc10764bd12 ). So manually started it. This issue is primarily discussing that problem. Namely a feedstock (repo) is freshly created and registered on Azure. Once we perform the first push of these two commits, Azure doesn’t appear to run.

However with this specific feedstock, it seems a recent PR ( https://github.com/conda-forge/ogre-feedstock/pull/2 ) is having the same issue. Not sure if that is a sign of the same underlying issue or not.

Should be able to point you to some fresh commits in the morning, which may prove more useful still.

Thanks for your help. 🙂

[leantk]

I would say try 120 seconds to play it safe if that works for you. This isn’t something I’ve seen before but it looks like for these 2 instances the definition creation time and the time the CI event came in are 30-40 seconds apart. We can manually kickoff builds immediately without a delay so the logic around CI events trying to find the definition must not be getting the most up-to-date information which I’ll dig into some more. For now, if you are ok with increasing the timeout, I think that gets you a workaround at least so you aren’t seeing builds not running every so often. Thanks letting us know about this and sorry for the trouble