Should a server be rebooted regularly??

[Tempest]

I've noticed that when my servers have been running for 100+ days, odd things sometimes start to happen. Like today one that was up for 142 days had the ftp stop working. Restarting the ftp server didn't help, had to reboot the entire server. So should a server be rebooted regularly? once a month? every couple months?

[liquidmoe]

Technically speaking unless you are performing some kind of maintenance that requires a reboot such as upgrading the OS you shouldnt have to reboot the machines. Some daemons may start acting funny after a while for different reasons, but you should be able to stop and restart them, freeing up all of their memory without an issue.

[detoxed]

I reboot mine every 5 minutes.

[directfiesta]

I reboot mine every 5 minutes.

.. and let it cool down for another five ....


Seriously : No.

And as stated above, you could once in a while NEED to restart some services, or kill background process ....

[fuzebox]

Any decent OS (read: FreeBSD) should never need rebooting short of a kernel upgrade... I've had servers go 2 years before a reboot.

You will have to restart services occasionally of course, due to memory leaks, runaway processes, etc.

[pornguy]

We try to have ours rebooted every 90 to 120 days. It only takes a few min, and the company that does it looks for slow traffic times to do it in.

[fetishblog]

The only time you should reboot your server is if you've upgraded the kernel or some other critical system component. Don't need to reboot it if you're just upgrading apache or proftpd or php. If something is not responding on it's own, you've got an issue you need to get checked out.

[Googled]

every minute or two. ;)

[flashbang]

up 267 days, 11:16, 2 users, load averages: 1.12, 1.21, 1.16

[react]

Unless, of course, you are running Windows servers. See eBay's weekly maintenance downtime, for example.

[Aric]

9:42PM up 546 days, 50 mins, 2 users, load averages: 0.12, 0.22, 0.18 .. and no problems

[minusonebit]

My NOC lost power and killed my uptime record of 700 days (bastards!)

Uptime 68 days 2 hours 36 minutes
Current Users 0
Load Averages 0.47 0.61 0.62

[Methodcash Rick]

top - 19:27:19 up 794 days, 23:02, 1 user, load average: 0.20, 0.43, 0.17
Tasks: 148 total, 1 running, 147 sleeping, 0 stopped, 0 zombie
Cpu0 : 3.1% us, 0.8% sy, 0.0% ni, 95.3% id, 0.1% wa, 0.1% hi, 0.6% si
Cpu1 : 0.5% us, 0.2% sy, 0.0% ni, 99.2% id, 0.1% wa, 0.0% hi, 0.0% si
Cpu2 : 0.1% us, 0.1% sy, 0.0% ni, 99.8% id, 0.0% wa, 0.0% hi, 0.0% si
Cpu3 : 0.1% us, 0.0% sy, 0.0% ni, 99.8% id, 0.0% wa, 0.0% hi, 0.0% si
Mem: 2074472k total, 1936188k used, 138284k free, 117072k buffers
Swap: 4096564k total, 144k used, 4096420k free, 1555248k cached

[u-Bob]

If your server needs regular reboots, something's wrong.