Tag Info

Hot answers tagged

6

It sounds like you may have two problems On your monitoring server, recording the metrics for lots of servers requires more random i/o than your storage can provide. Even if all your metrics are being written to disk, the server may be too overloaded to actually generate graphs from them. On your clients being monitored, the plugins which collect the ...


5

You can tell Monit to stop monitoring during certain times (ie while your backup is running). It uses the same syntax as cron. There are good examples in Monit Documentation in the Service Poll Times section. You probably want something like: check system example.com not every "* 2-4 * * *" This will stop monitoring every night from 2am-4am


5

Try strace -fF -o /tmp/debugfile your_script.pl And then grep open /tmp/debugfile If that does not catch everything, audit subsystem in Linux kernel could help you.


5

According to the docs, MRTG uses 1000 as the kilo multiplier by default. You can change it as follows: kilo[mem]: 1024


4

Is this disk cache? The operating system knows how to handle memory, and it usually does it good.


4

What's wrong with Cacti and Nagios? They're my preferences. You might take a look at Monit.


4

For one off debugging, you should use perfmon - it's included in windows. You can setup a counters to run in the background and collect data then analyze it later. Microsoft Explains it how just as good as anyone


4

If you only need to know if your disk is bottlenecking something, the basic parameters for disk performance are queue length (you might want to look at read and write queues separately, sometimes this is insightful), the disk idle time percentage and possibly the request service time. The queue length value tells you how many requests for disk data are ...


4

$ strace -f -e trace=process,open -o /tmp/script.strace script.pl


4

Absent full audit logging (every process run and the resources it consumed), you really can't. The best you can do is review all scheduled tasks (cron jobs, at jobs) and all the external influences you can catalog (scheduled jobs on other systems, an unusual request for a report coming from The Big Boss, etc.) to make an educated guess. The best way to ...


4

Honestly "Days Until Full" is really a lousy metric anyway -- filesystems get REALLY STUPID as they approach 100% utilization. I really recommend using the traditional 85%, 90%, 95% thresholds (warning, alarm, and critical you-really-need-to-fix-this-NOW, respectively) - this should give you lots of warning time on modern disks (let's say a 1TB drive: 85% of ...


3

Although being great tools, Munin and other RRDTool frontends (such as Cacti or Ganglia) have known i/o issues and are dificcult to scale when you monitor hundreads of nodes. There are some techniques to deal with this i/o bottleneck though. One of these thecniques is to spread writes across a large number of disks to reduce i/o in each disk. On the other ...


3

On Windows the only good way to find out (known to me) is to set an application pool for every site. By default everything is within a default application pool. Once they have their own application pool, they also have their own process. Having their own process will show you which one is taking all the ressources within the task manager. Good luck.


3

You just specify each file for the DEF: section of rrdtool. See below for a hacky one-off I did as an example. Notice that one of them is foo_kbrandt_foo1 and the other is foo_kbrandt_foo2. So the graph is pulling from two different rrd files. rrdtool graph MessagesDeliveredPerMinInfomationStores.png \ --imgformat=PNG \ --title="Messages Delivered Per ...


3

Yeah, check out /proc/$PID/stat, that should have what you need. The fields are defined in man proc. Can also pull memory usage and fault information, if you care about taht.


3

SonicWall/Dell solutions: SonicWall makes two product, ViewPoint (included in the Comprehensive Security bundle license) and Analyzer, that aren't too expensive (~$200USD), and will do exactly what you wish in a very easy to use way. If you don't want to invest, you may be licensed: Take a look here: ...


3

There are many tools available to help you track down issues like this. Start with the performance monitor which is provided with your OS. You can set it up to take periodic snapshot of pretty much any metric you can think of, including individual memory usage per process. In your case, that's probably the best tool for identifying rogue processes. Also ...


3

I contacted M/Monit support directly about the pricing issue... When asked if they offer a discount for education: "Unfortunately no, but for not-for profit organisations we might donate a license." But the fact remains that the unlimited license cost is intentionally low at ~$590. As for free alternative monitoring frameworks, I worked at a ...


2

I don't know of anything like this, but it seems like it would be relatively easy to handroll a nagios monitor to check for email in a mailbox and make sure that the last one that matched the specified subject line had happened no longer than X minutes ago. So if you expect to get 1 email/hr, make sure that there's always an email in the mailbox with the ...


2

This sounds like an application is leaking resources. Even good old task manager can help you see if that's the case: Go to the Processes tab. View menu -> Select Columns. Some columns to watch for abnormal values are Handle Cound, Thread Count (not very common) and GDI Objects. An application leaking handles or GDI Objects could very well cause what's ...


2

Iometer appears to be the overwhelming recommendation for testing and benchmarking iSCSI SAN performance. Because it is widely used by vendors, its numbers offer a fairly reliable comparison metric across storage systems. If you plan on running SQL Server off your SAN, you should also look at doing some testing and tuning with SQLIO. Brent Ozar has a great ...


2

Most people tend not to watch processor time but instead Processor Queue Length (number of threads waiting for a processor). That being said there's a good explanation here: http://www.windowsnetworking.com/articles%5Ftutorials/Key-Performance-Monitor-Counters.html


2

Take a look at http://www.centreon.com/ - That will wrap Nagios in a nice, user-friendly, GUI.


2

"best" tool is very subjective. Cacti & Nagios are both fine tools, each serving a different purpose (Cacti provides trending data, while Nagios provides real-time monitoring/alerting). I'm currently using InterMapper (commercial software, but reasonably-priced IMO), which does a little of both (though it's much stronger on the monitoring side than ...


2

it depends what sort of thing you are after for monitoring, i have been playing with groundworks at the moment (uses nagios as the backend monitoring system) but has all bee wrapped up in a web UI to make administration easier for people. Another good monitoring application i have been using for a couple of years is xymon (formally known as hobbit) this is ...


2

Check out Zabbix. It is one of the best Open Source performance monitoring tools out there. It scales well and has been used in environments with thousands of computers.


2

What I mostly do is just open a command prompt and then I type perfmon.msc. That gives me the performance monitor. Then I create a Data Collector set, with the counters I need. But after that analyzing the data collector log is the harder part. I wrote a blog article about analyzing performance logs on my own blog, it's here for your reference (and also a ...


2

Check out the Windows Performance Monitor. You can create a Data Collector Set that logs the metrics you want to a log file.


2

I don't believe there is any built-in system that does this, but nearly every monitoring system I've ever used is capable of trending disk space over time (at whatever granularity your polling cycle is, over varying & configurable lengths of time). If you do not have a monitoring system in place I would suggest checking out Cacti (if you just need ...



Only top voted, non community-wiki answers of a minimum length are eligible