How often or regulary are you rebooting your servers?

I got a raspy that is rebooting every 36 hours. Without those reboots, it used to hang up many times. How about you? Do you also reboot sometimes?


1 Like

I do occasionally trigger a reboot for updates (when I’m checking on the node every 4 weeks or so), but only because I run Windows nodes. I would assume Linux based systems will not need to be rebooted that often - most of my (non-Storj) Raspi projects have been running for months.

Every few months for kernel upgrades and/or due to power outage.

No scheduled reboots for me. System reboots are rare, every few months, I only do that for OS updates (from Synology). Node restarts only happen on node updates normally or when I want to change settings.

1 Like

My rpi4 dont really need rebooting they run fine for a few months on there own, I did have issues with sdcards causing me to reboot because of corruption and sdcards dying, Other then that I do not reboot them. Sometimes I need to shut them down for dust cleaning but thats when Ill do an update also.

1 Like

Same as @deathlessdd here. My RPi4 runs fine, and I rarely reboot it. Only when needed.

That suggests something is wrong. Once well configured, Linux systems are usually pretty stable and can run for years with no reboots. I’m not saying it’s necessarily a good idea, depending on updates that need to be run and so on, but… they usually do not reboot on their own like that.

Hmm, if I remember correctly, the hard drive made some problems. It was not readable anymore until I rebooted. Maybe a hard drive issue? It’s USB hard drive (HDD). Since I restart it automatically every 36h, it runs without problems. My uptime is 98,5 per Cent.

One issue you may have with rebooting your pi every 36 hours, is each time the node starts up it will run the filewalker process. If you are doing this every 1.5 days it will inevitably wear out your HDD sooner. And as your node grows, it will only get worse. It would definitely be best to figure out why this is happening and fix it.

I have a pi that I run a simple web server on that gets rebooted maybe once or twice a year. And my rock64 (pi like device) that I run my node on also never gets rebooted except for the odd power outage and maybe a software issue once in a while.

First, make sure all of your software is up to date. Then I would suggest you let your system hang and look in the logs for errors. You can find log files in /var/log/ and look in the syslog and possibly the kern.log files. You could also run dmesg from the command line and look there. There are many of us here who would be happy to help if you need it.

pi@storjpi:~ $ uptime
 14:45:28 up 264 days,  3:44,  1 user,  load average: 0.43, 0.45, 0.45
pi@storjpi:~ $ 

Can’t remember if my last power off was for some maintenance of my network closet or if it was a prolonged power outage where I shut it down before the UPS ran out.

1 Like

Ok guys, thanks for your participation on this issue. I now raised the reboot cycle to 300h. Let’s see how long it takes until it bugs again. Be patient! :wink:

Well I “try” to reboot every machine every 4 weeks so 28 days.
I run windows server based machines so they need the time you know update and everything.
but one of them has been happily working without a reboot for 158 days until a few days ago I did a reboot manually.

At my place of work we have a saying: reboot server once a week well at least once a month. If it has been running for longer than 3 months don’t even turn it off it will most likely not turn on

But a good rule of thumb is once a month for me

1 Like

I only reboot them, if there is a kernel update of medium or above severity kind of fix. Had servers running for over 6 months without reboot without any issues.

1 Like

I would suggest to use an external SSD and put the file system there. SDcard get corrupted fast.

Hi guys,
Short update: when I checked my node today it was still running, alive for more than 36h. That means everything alright. now restarting my node once every 300h. Thanks guys!

1 Like

Quite an improvement :slight_smile:
But maybe even that isn’t necessary, I would monitor the resources and see if it cannot stay up for longer, as rebrowing all files every 12-ish days is still quite heavy on disks.

But every 300h is definitely way better than every 36h :wink: :+1: