After migrating everything successfully in a machine to test, some data from influx seems to have been lost.
While running the migration command everything went fine but this error/warning came: 'schedule' from functions: Attribute already exists
Images:
Production server (how it should look after upgrade)
Test server after upgrade
Upgrade from 1.3.8 to 1.4.13. Both machines in ubuntu server
Summary
Title: [SOLVED] Upgrade issue: influxDB data loss and 'schedule' from functions: Attribute already exists
There was an upgrade from 1.3.8 to 1.4.13 on Ubuntu server. After the migration, some data from influxDB was lost and an error/warning "schedule from functions: Attribute already exists" appeared. The root cause is not fully identified, but it seems to be related to issues with the Ubuntu server. No solution provided in the thread.
D5
Moderator
Jan 3, 2024, 19:23
I think there was some time ago some similar issue with influx on cloud, where stats stoped working?
D5
Moderator
Jan 3, 2024, 19:28
Influxdb container logs:
Drake
Jan 3, 2024, 20:13
usage stats are all over the place though...
Drake
Jan 3, 2024, 20:14
cloud has a fix for the stats
D5
Moderator
Jan 3, 2024, 20:15
But why did a part just disappear after migration?
Drake
Jan 3, 2024, 20:53
usage data was sort of broken for a while on cloud..i wonder if that's related...i can't remember what the fix was
D5
Moderator
Jan 3, 2024, 20:58
No idea then, It's like the data just disappeared. Probably some part of my magic ability of crashing things involuntarily
D5
Moderator
Jan 3, 2024, 20:59
In fact for a moment it appeared right after console update but disappeared
D5
Moderator
Jan 3, 2024, 21:00
No idea. Anyways they're not very useful.
This error/warn: 'schedule' from functions: Attribute already exists
is something important or that I should fear?
Drake
Jan 3, 2024, 21:00
this is related to function scheduling for cron. It's probably fine
D5
Moderator
Jan 3, 2024, 21:01
Hope yes 🤞
Drake
Jan 3, 2024, 21:01
how old is this instance?
Drake
Jan 3, 2024, 21:02
what's in your telegraf logs?
D5
Moderator
Jan 3, 2024, 21:02
Created today and imported from backup
D5
Moderator
Jan 3, 2024, 21:02
If you mean the production instance, like 4 months
D5
Moderator
Jan 3, 2024, 21:02
Basically short prior to 1.3.8 release
D5
Moderator
Jan 4, 2024, 10:36
I don't see any errors there. Everything looks fine
Drake
Jan 4, 2024, 19:49
ok not the same as the cloud problem then..
Drake
Jan 4, 2024, 19:50
any errors in the usage worker?
D5
Moderator
Jan 4, 2024, 19:51
Nope, only aggregation success
D5
Moderator
Jan 4, 2024, 19:58
I'm closing this since It's not important at all.
D5
Moderator
Jan 8, 2024, 16:31
Seems the Ubuntu server wasn't working correctly and caused the issue. Marking this as closed since the root cause isn't 100% identified
D5
Moderator
Jan 8, 2024, 16:31
[SOLVED] Upgrade issue: influxDB data loss and 'schedule' from functions: Attribute already exists