We need to report our customer the time-based availability of their servers. Until now, we try to check it out in a complicate manner by analysing monitoring data. But isnt there an easy way to ask the server for its availability, kind of hidden log the in which the running keeps score of his own uptime?
Tek-Tips is the largest IT community on the Internet today!
Members share and learn making Tek-Tips Forums the best source of peer-reviewed technical information on the Internet!
-
Congratulations Chris Miller on being selected by the Tek-Tips community for having the most helpful posts in the forums last week. Way to Go!
Checker Server availability
- Thread starter stawolg
- Start date
- Thread starter
- #3
This will not help, because its pure uptime but related to what absolute timebase? This uptime must be related to time of service (normally 7x24h). And furthermore, planned downtimes via shutdown must be seperated from unplanned crashes. All in all, a combination of "uptime", shutdown/reboot events in syslog could be a way...
KenCunningham
Technical User
Another variable - just because your machine is 'up' doesn't mean the applications on it are necessarily available to users.
rzs0502
IS-IT--Management
- Nov 18, 2002
- 708
Maybe Nagios can help...
"If you always do what you've always done, you will always be where you've always been."
"If you always do what you've always done, you will always be where you've always been."
- Thread starter
- #6
Shure, we dont talk about availability of network (most important cause of outages i think) or external filespace. We only care about availability of server hardware and OS. I asked myself, why isnt this availability feature part of OS, because everybody wants to know about time of service. Like "check system time every second". If gap to last entry is greater than 1 second machine was down, check syslog entry for correct shutdown sequence, if there is none corresponding to last entry in time check table, uups, that was real downtime!
Should be elemtary part of OS, shouldnt it?
Should be elemtary part of OS, shouldnt it?
KenCunningham
Technical User
If you want an OS that spends most of it's time checking that it's alive and available rather than doing useful things, then yes.
- Thread starter
- #8
KenCunningham
Technical User
Got it in one. Shame they don't have a tongue in cheek emoticon on here, huh?
- Thread starter
- #10
Sorry dont understand the last post, but i feel that it will not help with our matter. Just to explain: the start point of this problem is not technical, its a matter of contracts made by controllers and managers who dont care about technical problems or performance, they just write down: 99,98% of availability of the service. And if it is than that more the provider has to pay. And if the provider is responsible for hardware and os he has to report about this, thats the business. Think nowaday reporting about your work is much more important than the work itself. Its stupid, but i cant help. Im just looking for a tool that can make reporting easier for me. It doesnt speed up work? So what, nobody cares...
KenCunningham
Technical User
I know, and I sympathise, but you have to be realistic about this even if the controllers and managers aren't. Did you look at nagios as described above?
- Status