Can I monitor log files for entries for a agentless server using piggy back
Can I monitor log files for entries for a agentless machine ??? I would like to monitor log files on a server that does NOT have a agent installed. Can I use the agent on say the Hyperic server to monitor the log file of a remote machine and set up alerts for a found entry. ??? I was looking at the file server file service but I dont see where I could specify a different machine from the one where this service would run ..
I really dont want to use NFS since there may be other files on different disks later that I might want to watch . Isnt there a way to look at (read files) on another computer that allows reading without using NFS . I know one can monitor using snmp devices that dont have clients. Now I just want to read a file and look for failures.
Let's back down to the physical on this. Is there any way to look at a file on another machine without, a) having some sort of app running on that remote host and displaying back to you, b) sharing the file/filesystem to a local machine where an app can read it? At some point, something on the remote host has to touch the file, either NFS, SMB, or hyperic. Something has to give you those 'eyes'. SNMP is an app running on the remote host that gives you those eyes.
The only workarounds beyond NFS, then, would be to set up a central log server and have your multiple hosts and/or apps send the logs there, and have hyperic on that log server (actually a very clean way to go if you monitor a lot of hosts, and very secure).
Or, write an app that runs on the remote host that can monitor the log file, then write a hyperic plugin that somehow queries that app and gives you the information you want? Or ssh to the remote and process/query/something? Either way I feel you will have to involve some intermediate app between the logs and Hyperic.
> I really dont want to use NFS since there may be other files on different > disks later that I might > want to watch . Isnt there a way to look at (read files) on another > computer that allows > reading without using NFS . I know one can monitor using snmp devices that > dont have clients. Now I just want to read a file and look for failures. >
Hey thanks for the reply.. I understand what you are saying . What Is actually happening. A network person has a web page setup so that all his switches and routers are monitored when one accesses the web page. I guess the web page has something setup that if its accessed ir runs a script and updates the web page with failed or OK for the network points. I also can reconfigure the web page to list the results in a comma separated format. I have created a python script that will access the page and create a file with the lines for each switch or router .on the local HYPERIC server. The question is now , this is a file that get updated whenever I run my script. Can hyperic and alerts be setup to watch this file even though it get regenerated each time so that I can wat the file for failures and have Hyperic alert people. The second question is can I get an alert sophistcated enouh to say if it fails 3 times in say 15 min send an alert. Since the file is regenerated and not appended IM wondering how Hyperic can count times it happened and where does Hyperic get its time stamp. From the log file or internall from its own alert ??
Also If I want to run this simple python script on my linux unbuntu 3.02 server every 5 min . Do I have to set that up with a cron or does Hyperic have a easier way to execute this python script that creates the results file(fail or ok).