using glance, or .. ?

[David_246]

Hi,

I am asked to find out a new monitoring tool with admin possibility's.
Now glance is a tool, that of course meets this requirement. But what tools do you use ? Are there some other (freeware) tools you use and think that can do the same or maybe even better ?
I already set up MRTG (cpu usage and load) and logsurfer that help me a lot.

Also they mentioned "advantage" (or something like that) in combination with glance. Now advantage is something I don't know. Has anyone of you experience with advantage ?

Any suggestion about (freeware) tools, experience with glance and advantage will be highly apreciated.

Best Regs David

[Ken Hubnik_2]

Sar and there is a tool called Big Brother but I have not used it.

[Paula J Frazer-Campbell]

Dave
A lot of use write tools to magage / monitor particular aspects of our server. That way you can tune it to do exactly what you require.

Paula

[Pete Randall]

Definitely a Glance person!


Pete

[Ken Hubnik_2]

Here is the link for Big Brother. www.bb4.com

[RAC_1]

There is NMM, then there is nagios. (www.nagios.org)

Nagios is free. BMC patrol etc.

[Tim D Fulford]

glance & MeasureWare.... without them we'd be lost. They are extensive and you can define & log user defined activities (e.g. you may have a database which has so many user threads, this can be logged too). Basically can't beatr them with a stick...

The other easy default alternatives are
top
ps
sar
iostat
who
+ losts & lots of scripts

my 0.02???

Tim

[Steven E. Protter]

I like glance but ....

There is a tool that comes with the ximian gnome environment that is pretty neat.

Its called the gnome system monitor.
It is a really souped up version of the top command with lots of extra features. An operations department can monitor a system pretty well with that.

If you want to monitor and control HP-UX and Linux systems, consider SCM.

Glance is the only tool mentioned above that costs anything but configuration time.

SEP

[Bill Douglass · Accepted solution]

I use glance for monitoring system utilization (memory, disk I/O rates, CPU). Top and vmstat are also helpful in this respect.

For 24/7 monitoring I use nagios (www.nagios.org). It's a bit of a pain to set up, but it does a wonderful job of monitoring disk space, critical processes, time sync,a nd many other items.

IDS/9000 is good for intrusion detection, and is available free from HP. Tripwire is also good, but is an extra cost product for HP-UX. I only use tripwire on my Linux servers, where it comes with the distribution.

MRTG is also good for monitoring traffic flow into and out of your server. This can be useful for long-term trend analysis, as well as trouble-shooting performance problems (e.q. why is the back-up taking so long.)

Finally, don't forget using sar to collect numbers over time. If you're not familliar with it, check out the sar and sa1 man pages.

[Caesar_3]

Hello!

The glance is very good, also MRTG.
Commands like : ps, top, vmstat, iostat,
swapinfo, sar.

Scripts that use the command and give reports.
Glance have language that give options to write
difrent reports.

Caesar

[Bill Hassell]

Another tool is SarCheck (www.sarcheck.com) which differs from Glance and top, etc as it provides recommendations rather than just raw data. For the casual sysadmin, this is a good choice and even provides HTML output for use as a web page. SarCheck uses sar, vmstat, iostat and other Unix commands to collect and analyze the system.

[Con O'Kelly]

Hi David

often the choice depends on your resources, budget & the size of the environment you're wanting to manage.

If you are managing 100's or 1000's of nodes and have the money and time to invest in the implementation then Network Management Software such as HP Openview Operations (OVO) and NNM are extremely powerful tools.
However for many sites they are overkill and they do require a significant investement in resources to implement properly and administer.

If you have a smaller site then I would advise, as I think Paula mentioned, writing your own scripts to monitor such things as FS utilisation, performance, key processes & security etc. It is fairly simple to write a few scripts that send all events to a global log file, which can then be displayed in a web browser. I have set it up in such way that all events are assigned a severity. If the severity is say "critical" it can post in the browser & also send notifications to pagers and emails.

Glance is an excellent tool. If you have a licence for GlancePlus Pak (which includes MWA) then look at using the /var/opt/perf/alarmdef file.
This is excellent for monitoring CPU, Memory, disk, FS utilisation etc and is extremely configurable to your particular systems requirements. It can be set up to email you or take other actions as soon as any thresholds are exceeded.

Cheers
Con

[Fragon]

Hi,David
From nagios's homepage, you can get many other monitor tools' link,here is the thread:

http://nagios.sourceforge.net/docs/1_0/about.html#requirements

Just at the bottom of the page!
-ux