DEV Community

Nick
Nick

Posted on • Edited on

What is your longest server uptime?

Hi all! My name is Nick and I work as a Full Stack Engineer.
Part of my daily job is configuring and maintaining servers!

I work in a very hybrid environment, consisting of instances in the AWS Cloud, Oracle cloud (very interesting experience that I hope to share in another post) and dedicated servers in our own and remote datacenters.

Today I migrated a server running a LAMP stack to a new location, and was shocked at it's uptime. It may not be that much but for me is my current record! This server survived working on an average load of 91% for 437 days!

serveruptime

So my question is, what is your longest server uptime? How did you achieve it?

Top comments (16)

Collapse
 
dmfay profile image
Dian Fay

SwiftOnSecurity on Twitter made the point ages ago that a year of uptime means that you've (hopefully!) applied a year's worth of patches, security fixes, and configuration changes but don't actually know if the system can come back up if+when it does reboot for any reason in your control or outside it. It's stuck with me.

Collapse
 
nicolasguzca profile image
Nick

Very good point! I must confess this server had all mayor updates except the Kernell ones.

Collapse
 
moopet profile image
Ben Sinclair

I don't know exactly but probably about a decade.

I worked for a while on recovering an application from a running server which hadn't ever been turned off and had "started making funny noises". I had to telnet to a SunOS box and grab the code and try to replicate functionality on a modern (well, modern for a few years ago) SUSE box. It wasn't just a case of updating the code, we weren't sure exactly how the system was supposed to work so I needed to run it in parallel on both systems to make sure my modifications weren't breaking functionality. That was fun :)

Collapse
 
nicolasguzca profile image
Nick

That is some Sensei level stuff

Collapse
 
mrmainnet profile image
Nguyen Tan Vy

You should try kernelcare.com/ for kernel update without reboot.

Collapse
 
nicolasguzca profile image
Nick

Awesome, will definitely check it out

Collapse
 
michaelgv profile image
Mike

Where I work, we have a server thatā€™s been up for 7 years, itā€™s a 32bit database server, not public, on its own vlan, itā€™s quite a beast - itā€™s officially EOL for us now, so weā€™re migrating it to our new 64 GB RAM, 8x2T server) (full disclosure: weā€™re our own data center)

Collapse
 
nicolasguzca profile image
Nick

Wow 7 years? That is a lot!

Awesome server you are migrating to. Our SQLServer main server has similar specs but in a cloud provider.

Collapse
 
nebojsac profile image
Nick Cinger

I'm working on retiring a server now that has an uptime of 1070 days, so about 3 years!

Was happy to see it past the 1000 day mark, ha!

Collapse
 
nicolasguzca profile image
Nick

Awesome! 1000 days is a lot!

Collapse
 
nebojsac profile image
Nick Cinger

Yeaaaaa, it's more of a thing to worry about, than a thing to brag about, as some have said haha.

That's not to say that nginx or mysql haven't "gone away" several times in the past 3 years :)

Collapse
 
michaelgv profile image
Mike

My longest server uptime is 7 years, the box was a database server, internal only. It was taken to the trash today. All cloud means all fun.

Some comments may only be visible to logged-in visitors. Sign in to view all comments.