@Jon I’ve reset the PG database, so that everything is fresh, the DB was created on init and the PG OS Timezone is Europe/London , and everything was setup from the postgres startup, some interesting things happen while running that I didn’t spot before.
I imported my credentials and imported just ONE of my workflows to show the DB timezone info etc
and I was looking at the execution details when I noticed this happen, when the execute starts running it showed local time, then when it completed it switched back to GMT…
Not had chance to try the postgres db on my RPi Debian platform yet… BUT
I’ve noticed … Subflow executions are showing the correct Timestamp ! While the calling Main Parent work flow shows does the strange Flip back to GMT when complete.
There seems to be some issue around the use of the GENERIC_TIMEZONE when having O/S TZ environment set to be the same. I reproduced the issue on my own Debian environment.
If I keep the container running TZ=UTC and only set N8N’s GENERIC_TIMEZONE=Europe/London then everything appears to be good, schedules execute at the right local time, and execution details are displayed correction, webhook timeouts play nice… etc… even the DB can be in UTC as well which keeps a few people happy.
That is useful to know, oddly enough all of my options are Europe/London so there is still a bit of a question as to what is going on.
I may have to set aside some time to have a play on a clean environment, maybe when the clocks change again so I don’t need to worry about manually messing with the time.
hey, @0101binary0101 your workflow setup to update your n8n version using n8n it’s amazing. Would you mind sharing the workflows? I’m really interested in automating that part of the work.