February 28th, 2006, 03:54 AM
I am wanting to find the BEST (relative) tools to accomplish a multitude of monitoring. I would prefer as few tools as possible. I want to keep expenses down to below a few grand. I need to monitor multiple different things.
Environment: High production with 24/7 uptime required.
Windows servers (mostly windows 2k3, 1/3 windows 2k)
Type of servers:
Compaq/HP Proliant servers (from older dl360's 5000's to month old HP's)
What needs monitored: Almost everything and if possible metrics provided with it
Monitor event logs(I dont want to go through 30 servers worth of logs either I would prefer a tool that will auto check for specific things and then email me)
Montior Hardware (hard drive, power supply etc fails)
Monitor if a web page goes down
Monitor any services uptime/downtime
Monitor local applications/batch files
Monitor SQL server
Monitor if certain files are changed
Anything another server admin might find important to them
How to contact: Via email to my cell phone/pager/blackberry
Currently we have and own What's up Gold and some other apps but they don't allow for specific services or homemade services or a lot of the other things I want to get updates on.
I am looking for the expertise of others and hopefully this will turn into a growing thread to help anyone that gets into this field as well.
Thanks in advance
Duct tape.....A whole lot of Duct Tape
Spyware/Adaware problem click
February 28th, 2006, 04:22 AM
I have investigated only a few smaller tools...
I have used RRDtool which is highly customisable, (but obviously the admin overhead would be high)
I have also used BigBrother and nmap as well... which both require heavy administration, at least to set up.
Nagios has been recommended to me in the past.
I quickly googled and came up with this link. It has many many network monitoring tools for many different requirement sets.
Hope you are successful culling the network monitoring tools. Can you let me know which one you decide upon?
February 28th, 2006, 04:23 AM
/disclaimer: I dont work for Mercury nor being paid by them....just love the product.
I use a product called Sitescope at work. It's made by Mercury Interactive. It's a web applicaiton that runs on Windows and provides for many different monitors such as URL, URL sequence, SQL, port, DNS, email, Windows services, Windows processes, disk space, will run scripts (custom made by you or stock), Windows perf counters, etc etc etc. Over 65 monitors - see here http://www.mercury.com/us/products/b...supported.html
Other nice things:
* Doesnt require an agent. Will work with different authentication schemes/AD domains etc
* Alerting is very flexible
* Monitors have lots of options to tweak if you want
* Reporting is very good allowing you to report on uptime, response times, etc during user specified reports periods and even allows you to tell it during what times during those days. Just a lot of flexibility.
It sounds like something you are looking for. We love it.
One word of caution: it's EXPENSIVE. The licensing is based on monitor points and most monitors only use one point each (URL sequence uses 1 point per step). The points are $72+yearly maintenance each so if you want to monitor a particular Windows server for example you may be looking at needing say 12 points (for PING, disk space for 3 drives, services for 5 services, couple ports, etc)...which if you have lots of servers can be costly.
More info at: http://www.mercury.com/us/products/b...ter/sitescope/
They offer a 10-day demo.
Hope this helps. If any questions about it feel free to ask as I know the product very well....it's our primary monitoring app.
February 28th, 2006, 09:44 AM
Experience is something you don't get until just after you need it.
February 28th, 2006, 12:27 PM
In terms of services, we use Nagios , which is an open source package similar to WhatsUp Gold. It comes with a bunch of stuff for common services, and we've written a few routines to handle particular apps, our Internet download limit etc.
For event logs, one of our contractors left us a script which rips through the event logs on a server nightly, ignores anything you've told it to expect and then sends an email with anything unusual. Unfortunately they consider it part of their service; I can't forward it.