| NeilHanlon | fungi: I can confirm that the endpoint I gave to mnasiadka can be used for opendev to pull from | 02:17 |
|---|---|---|
| NeilHanlon | I think that there is definitely some value of jobs which could be improved by having a rocky mirror. in particular, whenever rocky might release something and it takes some time to propagate to all the mirrors, there can be spurious CI failures in both kolla and OSA due to mismatched content on mirrors. At least with an opendev mirror, the | 02:19 |
| NeilHanlon | artifacts should be consistent | 02:19 |
| Clark[m] | They would only be as consistent as what we sync from so that depends on the endpoint | 04:06 |
| *** maxh1 is now known as maxh | 07:47 | |
| *** maxh1 is now known as maxh | 07:56 | |
| fungi | there would need to be some consistency checker we could run over what we've rsynced in order for the script to be able to avoid publishing an inconsistent state it retrieved from the source site | 13:21 |
| fungi | otherwise, yes, if the rsync source is in an inconsistent state then so will our mirrors be | 13:21 |
| *** mtreinish_ is now known as mtreinish | 18:47 | |
Generated by irclog2html.py 2.17.3 by Marius Gedminas - find it at https://mg.pov.lt/irclog2html/!