Hi, I only just installed ROCKSTOR a couple of days ago, configured a couple of drives as RAID1 and copied some data over, oh and installed Plex. This has happened a couple of times, but basically it is not working at all, cannot logon to the web, plex is offline, the drives were unmounted. Rebooting brought the drives back online, but still cannot logon to web. SSH is fine.
Error log is here:
http://1drv.ms/1Pp43fK
Thanks
Thanks for the logs. The error is clear and coincidentally, I saw the same problem on a machine here this morning. The problem is that postgres connections are accumulating over time causing this error. However, I am not sure of the root cause yet. Looking into it.
When you see this problem, you don’t need to reboot. Just restart postgres: systemctl restart postgresql.
I bet this is the same issue I reported in Fatal error: Too many clients error when connecting with https after a while. After a while, I see the same thing, i can’t use the UI, the ui pages are all page not found.
While I am thinking of it, what is the extensibility model for using postgres - I assume I can count on it running, and if I want to create tables and store data there, that’s fine, I should just make sure I back up my tables and schema before a system update or something. I searched in your docs and only found this reference, http://rockstor.com/docs/contribute.html?highlight=postgresql, which seems to suggest I could create my own database and use it.
Hi @Nick, yes it is the same issue. I haven’t gotten to the root cause yet, seems like there’s a slow connection leak perhaps triggered by data-collector service. Are you able to consistently reproduce it? If so, are you able to reproduce it even if you leave the UI in static page such as Storage -> Disks?
Rockstor creates two databases during first install and these databases are never deleted. Any database other than Rockstor’s own are not messed with either. You sure can create your own.