Re: Scheduling Weekly Reboots
Quote:
Originally Posted by
TheFu
Hope you realize those are just comments for reminders what each column means. Nothing more.
I do and thanks (seriously). The service should restart on Sunday; I'll check and see how responsive the server is after that and report back. I copied that data and put it in the crontab file for future reference.
Thanks again!
Re: Scheduling Weekly Reboots
Quote:
Originally Posted by
linuxnoob87
I know that if I don't reboot weekly (or so) that I can't SSH in. It's running in a VM with plenty of RAM and CPU. I'm hoping that weekly restarts of the Plex service will fix the issue.
something is wrong somewhere if it doesn't let you SSH after any number of days of uptime. most everyone else here knows that and would want to figure out what it is that is wrong. it could be a misconfiguration in plex or somewhere else.
have you considered the idea of just restarting SSH every week instead of the whole system?
Re: Scheduling Weekly Reboots
I run a plex server on 16.04; upgraded it from 14.04 before that. Never seen anything like this. I doubt restarting plex will help. Check the system log files for any warnings or errrors. Look at the nearby entries to see what is giving issues. A failing disk with corrupt data can cause all sorts of funky issues. So can a not-yet-dead PSU.
ssh failures would be a network issue or firewall issue, IME. I would turn up the logging for all clients and on the server to solve that issue.
If ssh crashed, I'd physically go to the machine and login there, trying different TTYs until one worked. If none worked, I'd look at systemd-logind for the issue.
Startup stuff has changed with every recent LTS. We don't know the release of the OS being run.
But that wasn't the question in this thread.
Re: Scheduling Weekly Reboots
Quote:
Originally Posted by
Skaperen
something is wrong somewhere if it doesn't let you SSH after any number of days of uptime. most everyone else here knows that and would want to figure out what it is that is wrong. it could be a misconfiguration in plex or somewhere else.
have you considered the idea of just restarting SSH every week instead of the whole system?
Would it make more sense to restart SSH or the Plex service weekly?
I've not done much to my install of Ubuntu Server so I don't know why this happens (just that it happens)
Re: Scheduling Weekly Reboots
Quote:
Originally Posted by
linuxnoob87
Would it make more sense to restart SSH or the Plex service weekly?
I've not done much to my install of Ubuntu Server so I don't know why this happens (just that it happens)
No. Restarting services that run fine, indefinitely, for everyone else in the world isn't a solution. Linux isn't Windows.
Figure out the root cause, fix that.
Re: Scheduling Weekly Reboots
Quote:
Originally Posted by
TheFu
No. Restarting services that run fine, indefinitely, for everyone else in the world isn't a solution. Linux isn't Windows.
Figure out the root cause, fix that.
Ok. How?
Linux (specifically Ubuntu) is still very new to me so I've no idea how I'd even begin to troubleshoot this issue. Not trying to be snarky; I legit don't know how I would troubleshoot this.
Re: Scheduling Weekly Reboots
Quote:
Originally Posted by
TheFu
I run a plex server on 16.04; upgraded it from 14.04 before that. Never seen anything like this. I doubt restarting plex will help. Check the system log files for any warnings or errrors. Look at the nearby entries to see what is giving issues. A failing disk with corrupt data can cause all sorts of funky issues. So can a not-yet-dead PSU.
ssh failures would be a network issue or firewall issue, IME. I would turn up the logging for all clients and on the server to solve that issue.
If ssh crashed, I'd physically go to the machine and login there, trying different TTYs until one worked. If none worked, I'd look at systemd-logind for the issue.
Startup stuff has changed with every recent LTS. We don't know the release of the OS being run.
But that wasn't the question in this thread.
Running a Plex server on 16.04 as well. What logs would I be looking at (specifically where would they be and what would I be looking for)? The install is running on an ESXi VM so I don't think it's the HDD and the server has redundant power supplies so I don't think it's that either.
The firewall is disabled on the VM. How do I "turn up the logging"?
Re: Scheduling Weekly Reboots
Quote:
Originally Posted by
linuxnoob87
Running a Plex server on 16.04 as well. What logs would I be looking at (specifically where would they be and what would I be looking for)? The install is running on an ESXi VM so I don't think it's the HDD and the server has redundant power supplies so I don't think it's that either.
The firewall is disabled on the VM. How do I "turn up the logging"?
On all Unix-like systems, logs are almost always somewhere under /var/log/.
Look for warnings and errors. I would use egrep to search all of them quickly.
Being a VM doesn't prevent disk failures.
How to turn up logging is specific to each program. You'll need to read the manpage for each, find the section on logging, debugging, or verbosity settings, then do what it says. For the sshd (server), I think it requires running it manually with the --debug, but I didn't check that. You should.
For ssh clients, getting more details is enabled by adding more 'v's to the connection request.
Code:
$ ssh -vvv userid@remote-server
more v's gets more details. I've pretty certain about this, but you should check the manpage. Never trust options from people posting on the internet. The difference between a -p and a -P can be huge.
But since this really isn't the same question as the original, a new thread is needed to help the community.
If you feel you know how to schedule automatic reboots now, please use the "thread tools" button near the top and mark this thread as SOLVED - which helps people seeking answers AND helps stop people looking to help from wasting time on already SOLVED problems.