floral gate
Not sure how to recover from this apart from last night's backups?
I followed the instructions as defined here (https://twenty.com/developers/section/self-hosting/upgrade-guide - as I always do without issue) and then today this happened.
yarn database:migrate:prod
yarn command:prod upgrade
The migrate:prod command seem to run without issue. The prod upgrade is the problem?
thorn solstice
Hello @floral gate, what version of twenty are you running exactly please ?
Seems like your workspaces failed the migration
Do you have several workspaces or just one ?
floral gate
Just one workspace, was trying to upgrade from v0.42.18 to the latest which is v0.43.2
Thanks
thorn solstice
Have you tried re-running the upgrade command ? Might be a cache warm up issue
What error do you have in your browser console exactly ?
floral gate
I have not tried running it again yet, will try now.
Okay I just did the following and the local storage is empty?
/app/.local-storage $ ls -al /app/.local-storage
total 12
drwxr-xr-x 1 node root 4096 Mar 3 11:23 .
drwxr-xr-x 1 node root 4096 Mar 3 15:19 ..
/app/.local-storage $
I mean if anything is going to cause it to fail, that would do it right? I'm sure there is supposed to be a folder with the workspace identifier etc...
thorn solstice
please also run
yarn command:prod upgrade
at least twice in a row
Just to attest this is not due to some caching race condition
floral gate
Ran twice in a row
thorn solstice
Ok so indeed there's a problem
Have you made manual/specific update to the core or metadata tables lately ?
floral gate
No not at all.
thorn solstice
When would you be up for a call ?
floral gate
Yeah sure thing, that would be great.
thorn solstice
I can be available in half an hour right now if you're too we can jump on this
floral gate
I'll need to get ready to leave in 30 minutes, happy to talk tomorrow if that is better for you?
thorn solstice
Is your redis instance up and healthy ?
floral gate
Yes
thorn solstice
Hey @rigid frigate @lyric violet, do you guys have please heard of such issue in the latest 0.43 with the dataSourceMetadata or cachedObjectMetadataMaps retrieval along the upgrade command ?
Just in case@floral gate please restart your redis instance nor wipe it all and retry the upgrade command ( twice 😛 )
floral gate
Will do.
rigid frigate
Hey <@1325771357702721557> <@1156119986058834012>, do you guys have please heard...
unfortunately, not heard of any issue with the dataSourceMetadata or cachedObjectMetadataMaps
thorn solstice
@floral gate I've just pushed a 0.43.3 that might fix your problem, please let me know
At least you would get other errors 🙃
floral gate
Okay thanks I'll try it now. But I'll have to head out in a little while. Let me try what you have suggested, otherwise I'll be continuing tomorrow.
thorn solstice
No worries no hurry
floral gate
I can't see it here:
https://hub.docker.com/r/twentycrm/twenty/tags?name=0.43.3
I'm working with the docker images on Kubernetes.
I do see the source https://github.com/twentyhq/twenty/tree/v0.43.3
floral gate
Great thanks, I might have a moment this evening just to hop on and see it makes a difference.