Messages in this thread |  | | Date | Thu, 6 Feb 1997 17:18:41 -0500 (EST) | From | David Schwartz <> | Subject | Re: reliability |
| |
On Thu, 6 Feb 1997, Richard B. Johnson wrote:
> Linux reliability has been questioned. Here is an example of demonstrated > reliability. I re-booted it a few days ago to replace one of the tape drives > which has worn out and needs to be refurbished.
We have a Linux machine here -- one of our heavily-loaded name servers handling primary/secondary DNS for over 100 domains and it's used for name resolution by over 100 machines. Some of WIZlink's customers also use it for backup mail handling. It also runs a large IRC hub and a web (Apache) server that gets well over 10,000 hits a day. It was last rebooted to upgrade to 2.0.27. It's a P133 clone with 32Mb of RAM.
knock:~$ cat /proc/version Linux version 2.0.27 (davids@davids) (gcc version 2.7.2.1) #1 Sat Dec 7 15:04:37 EST 1996 knock:~$ uname -a Linux knock 2.0.27 #1 Sat Dec 7 15:04:37 EST 1996 i586 knock:~$ uptime 5:15pm up 56 days, 23:48, 1 user, load average: 0.27, 0.24, 0.18
It has exactly one problem, it gets these "TCP: **bug**: copy=0, sk->mss=0" a couple of times a week. No other problems have been observed.
David Schwartz
|  |