Best Linux Server uptime?

[Red Squirrel]

Online web server still chugging along:

[root@shared2 zones]# uptime
20:34:58 up 1087 days, 53 min, 1 user, load average: 1.85, 1.76, 1.73
[root@shared2 zones]#

Had to shut down all my home stuff yesterday due to some power work (moving UPS and switching to new battery plant setup)

So that one is now back to square one. It was at around 100 before I had to turn it off.

[root@borg ~]# uptime
20:38:41 up 23:24, 2 users, load average: 0.37, 0.39, 0.42
[root@borg ~]#

That server always gives me trouble when it has to be rebooted, so I'm hoping this is the last time ever I have to shut it down. Batteries are good for 4 hours but after 2 I'd probably be shutting down other stuff to make it last longer. Eventually I just want to get a generator.

 

[Red Squirrel]

Bump.

At some point I switched my web server to another provider, so it's uptime is now sitting at 452 days. I recall the uptime of old server reaching well over 1400 days. I eventually lost that uptime when they had a UPS failure. That's not why I switched providers though, I tend to move providers or just new servers within same provider once in a while as it comes a point when I can get something better for same monthly price so may as well do it.

For home servers:

VM server, Moria is at 583 days

File server, Isengard is at 585 days

Main (first ever) server, Borg is at 581. I am actually working on retiring that box to a new VM. (new OS installation, not just a P2V) It currently only does email, used to do DNS, file, web apps, etc...

My home automation server, Hal9000 is at only 192 days. At one point my relay controller hosed up and it was just easier to do a reboot to reset it as I was not physically home and could not tell the temperature. Got an alert on my phone that it had locked up.

Firewall is at only 37 days, it rebooted on it's own the other day, really not sure why, but I have my fingers crossed that it's an isolated incident. I think it may be a packet capture that I forgot about, which eventually used all the memory.

I eventually need to reboot my file server though to see if a kernel update fixes a random file I/O related crash, but that means rebooting everything else too, so I'm kinda weary about it and just have not gotten the courage to do it yet. But eventually I'll have to.

 

[John Connor]

Wow that's pretty good for a netbook! Being a "workstation" it's much more vulnerable to being rebooted given it's being used directly.

I don't think I've ever gone more than a week with any of my workstations just because I'll usually end up having to reboot for a reason or the other. Updates, or whatever.

I have a netbook in the kitchen I call the kitchen kiosk that serves a FTP server and Teamspeak server that's been on 24/7 for 2 + years now. LOL. It's a Dell mini 910 running XP SP3.

Got all sorts of iptables in the router using DD-WRT, I use non-standard ports, and have other security measures. Thinking about getting another. I clone the thing from time to time on a USB stick and it only takes about 15 minutes. LOL. Oh! The FTP server stores on the internal SD card. LOL

 

[Maximilian]

05:30:38 up 20 days, 16:56, 3 users, load average: 0.11, 0.06, 0.05

Debian 8.2

It wants to restart now and again after updates, doubt it'll ever make a years uptime :\

 

[ultimatebob]

I've had some Linux and AIX boxes with over a year of uptime before I had to reboot them for a scheduled power outage or a kernel patch.

My Windows servers can usually only go a month or two before needing a reboot for a security patch.

 

[dave_the_nerd]

We have one VM host that's at 695 days.

 

[Red Squirrel]

Just realized one of my servers is at 824 days. The one that I've been suppose to retire, but still chugging away handling email.

[root@borg ~]# uptime
02:15:00 up 824 days, 8:42, 1 user, load average: 0.07, 0.02, 0.00

It's a core2quad running Fedora Core 9. Updates, what's that? Don't worry that server is not internet facing.

My other servers have low uptimes now due to an incident where my file server went down due to UPS not tripping to battery fast enough during a brown out and that involved rebooting everything else that relied on it. The other servers were unaffected for some reason despite being on the same UPS. It's a mystery what happened there. I'm just glad I did not lose any data, only had one database that got corrupted with all that so had to restore it from backup.

 

[Fallen Kell]

We just ran across a system which had an uptime of 6221 days at my work. We did some quick math and realized its last reboot was most likely to patch the kernel for Y2K bugs.

 

[Chaotic42]

As of right now in the house?

12:34:09 up 2 days, 4:31, 2 users, load average: 0.09, 0.12, 0.09 :beer:

 

[Red Squirrel]

We just ran across a system which had an uptime of 6221 days at my work. We did some quick math and realized its last reboot was most likely to patch the kernel for Y2K bugs.

Oh wow lol, that's quite nuts.

 

[JM Aggie08]

I've moved to a new role within my company, but when I did support the unix/linux environment, the highest I saw was ~900 days...which is heresy at this company.

 

[sourceninja]

...

 

[Red Squirrel]

My oldest home server is sitting at 1052 days now. It's only doing mail, I eventually want to migrate it to a VM and retire that box. The OS installation on that box is like over 8 years old, but it just keeps chugging along.

 

[Red Squirrel]

Not my best but had one VM on almost 1 year, but I broke that. Wanted to add more disk space to it so had to restart it. You'd think Vmware could just hot add it though. Probably have to pay extra for that.

[root@aovdb aov]# reboot

Broadcast message from root@aovdb.loc
	(/dev/pts/4) at 20:30 ...

The system is going down for reboot NOW!
[root@aovdb aov]# uptime
 20:30:59 up 332 days,  4:40,  3 users,  load average: 0.66, 0.21, 0.12
[root@aovdb aov]# Connection to aovdb.loc closed by remote host.
Connection to aovdb.loc closed.
ryan@falcon ~/temp $

 

[Fallen Kell]

Not my best but had one VM on almost 1 year, but I broke that. Wanted to add more disk space to it so had to restart it. You'd think Vmware could just hot add it though. Probably have to pay extra for that.

[root@aovdb aov]# reboot

Broadcast message from root@aovdb.loc
    (/dev/pts/4) at 20:30 ...

The system is going down for reboot NOW!
[root@aovdb aov]# uptime
 20:30:59 up 332 days,  4:40,  3 users,  load average: 0.66, 0.21, 0.12
[root@aovdb aov]# Connection to aovdb.loc closed by remote host.
Connection to aovdb.loc closed.
ryan@falcon ~/temp $

You can add a disk live with workstation, pretty sure it would work with player (but havn't tried it):
https://www.cyberciti.biz/tips/vmware-add-a-new-hard-disk-without-rebooting-guest.html

 

[Red Squirrel]

You can add a disk live with workstation, pretty sure it would work with player (but havn't tried it):
https://www.cyberciti.biz/tips/vmware-add-a-new-hard-disk-without-rebooting-guest.html

This is in ESXi. I was able to add it live, but it only applied it after a reboot. Ex: when I did dmesg it did not show as a disk being added.

 

[dave_the_nerd]

This is in ESXi. I was able to add it live, but it only applied it after a reboot. Ex: when I did dmesg it did not show as a disk being added.

Did you add a second vmdk, or just expand an existing one?

You know about this, right?

 [root@redhat ~]# echo "- - -" > /sys/class/scsi_host/<host_number>/scan

 

[JM Aggie08]

Did you add a second vmdk, or just expand an existing one?

You know about this, right?

 [root@redhat ~]# echo "- - -" > /sys/class/scsi_host/<host_number>/scan

This.

 

[Paperlantern]

Wow, this thread still getting replies. Ive since retired the box I posted about in the OP, don't play much MineCraft anymore these days. Or much PC games in general. More into retro games on consoles that are over a decade old. NES, SNES, N64, GameCube and Wii. Fun collecting for those systems now, hard to find stuff. Hate the newer consoles. Switch is currently a joke in my opinion. But enough of that.

I don't actually have any Linux servers at home anymore. Some of the ones at work are only up for a few months tops, so nothing impressive here anymore. The one up for over 6000 days blew my mind!

 

[wirelessenabled]

We still reboot our HP-UX servers every night with a cron job. Having said that we have an old D Series server that has been running since about 1996! It has a legacy application on it that will never be ported to something more modern but just can't seem to be eliminated.

 

[dave_the_nerd]

We still reboot our HP-UX servers every night with a cron job. Having said that we have an old D Series server that has been running since about 1996! It has a legacy application on it that will never be ported to something more modern but just can't seem to be eliminated.

Your servers are old enough to drink?

 

[Red Squirrel]

Did you add a second vmdk, or just expand an existing one?

You know about this, right?

 [root@redhat ~]# echo "- - -" > /sys/class/scsi_host/<host_number>/scan

Was a new separate VMDK, figured I'd keep data separate from OS.

What does that command do, I know it's writing to that particular virtual file, but what exactly does that do?

 

[dave_the_nerd]

Was a new separate VMDK, figured I'd keep data separate from OS.

What does that command do, I know it's writing to that particular virtual file, but what exactly does that do?

Forces the host to rescan the scsi bus, basically. So it'll make a new drive show up w/o rebooting.

<host_number> can be one of a couple different scsi hosts. Do them all.

 

[Red Squirrel]

Oh that's good to know, I'll have to try to remember that. (probably won't - not really something I run into often )

 

[dave_the_nerd]

Oh that's good to know, I'll have to try to remember that. (probably won't - not really something I run into often )

I just did it again this morning. (Resizing a LUN on a file server from 32 to 48TB.)

Wheee!