Great Uptime FYI

A

alvinswim

Just wanted to share with you guys out there.. We have a host that's been online for 1774 days.. it's an internal loghost but it's done very well over the years... and definitely has out-lived its own hardware... as well as the rest of our infrastructure!

Code: 
[axxxxx@xx-xxxx-log-1 axxxxx]$uname -a
FreeBSD xx-xxxx-log-1.xx-xxxx.xxxxxxx.net 5.4-RELEASE FreeBSD 5.4-RELEASE #0: Sun May  8 10:21:06 UTC 2005     
[email]root@harlow.cse.buffalo.edu[/email]:/usr/obj/usr/src/sys/GENERIC  i386

[axxxxx@xx-xxxx-log-1 axxxxx]$uptime
 4:28PM  up 1774 days, 51 secs, 2 users, load averages: 0.23, 0.15, 0.17

It's a trooper!
 
Uptimes are overrated.

As Alt noted, it's highly vulnerable to all security issues that have been published since this thing went online. Definitely not something I would be proud off and showoff to the world.

It might not be connected to the big bad internet but you have much more to fear from your own employees. They don't need to hack into your network, you already gave them access. They usually also know which servers are the important ones.
 
It is really amazing to think that a machine has been up constantly doing what is supposed to do for 5 years.
On the other hand, there are many security risks involved when you are not patching the OS.

A good compromise can give you a year of safe uptime!

Regards
 
This is surely a great test about how long a FreeBSD can stay online.
Apart this, I would suggest an upgrade as other pointed out. Besides the security risks of an old system, there is also a hardware checkup to be done (even if the machine is not business critical), 1700+ days are a lot even for good hardware.
 
well luckily for us this guy is on a segmented network where only myself and my colleague have access to and only syslog for network equipment go to it.. its pretty safe from the outside..
 
Well, granted it is IP based uptime, there could, and should be failover in place, but there still very well could be some single machines in that mix.

Slightly offtopic, but what bothers me more about that list, is there is no BSD listed, 6 years ago, it was ruled by BSD, and Linux machines.
 
Dru said:
Slightly offtopic, but what bothers me more about that list, is there is no BSD listed, 6 years ago, it was ruled by BSD, and Linux machines.
Click to expand...

Don't be.

Think for a minute, I have a Windows 2003 server with a few years of uptime. The server is connected to the Internet. Am so proud about this that I post the IP of that server online.
 
Nice uptime.

But from personal experience, we had a host with just about 3 years uptime, which we powered off to relocate to a new datacenter. Sadly, it never came back up and as it was a semi-critical system, we had to scramble to find a replacement for it, this was already on top of a long delay with the relocation of the datacenter.

Now uptime isn't as important, so long as its not between 1 - 30 days ;-).
 
gkontos said:
It is really amazing to think that a machine has been up constantly doing what is supposed to do for 5 years.
On the other hand, there are many security risks involved when you are not patching the OS.

A good compromise can give you a year of safe uptime!
Click to expand...
That actually makes me wonder about something. In the aerospace industry there are basically three design philosophies regarding failure mode: "failsafe", "safe life" and "damage tolerance". Does the computer industry (and more to the point: FreeBSD) use similar concepts for relating maintenance to reliability?

Fonz
 
You must log in or register to reply here.
Back
Top