Tom Parrott
Tom Parrott
Thanks @niemeyer. Yes I would appreciate your thoughts on this too, we're not going to make a final decision before Madrid as we want to ensure the other `micro*` teams...
Please can we chat about this next week at the sprint. This is starting to sound rather complicated and i dont understand why it has gotten to be this way....
Also LXD has been known to have to do things like this occasionally in patches, but to my knowledge its not done routinely like this.
Thanks Max. I think this warrants an architectural discussion around what micro cluster provides to its consumers. Thanks for laying out the landscape of options.
> * If we detect that some nodes are behind with the `schema.ErrGracefulAbort`, we return early. Is this effectively the same as `patchDBNodesAutoInc` in LXD? Albeit it sounds more complex...
@cole-miller do you think you will have time to make the suggestions from @ru-fu so we can merge this? Thanks!
@cole-miller thanks, although the markdown linter is unhappy: > .tmp/doc/howto/dqlite_troubleshoot.md:4: MD009 Trailing spaces
Also please can you rebase your commits so that the commits in this PR tell the final version of your contribution, rather than showing the evolution of it.
Has the idea of passing an FD been proved to work? Before we get too bogged down in code style etc?
@hamistao once this is ready for review please can you mark it as ready for review and let me know. Thanks