[ZBXNEXT-231] Current Disk Queue Length Created: 2010 Feb 12 Updated: 2024 Jul 24 |
|
Status: | Open |
Project: | ZABBIX FEATURE REQUESTS |
Component/s: | Agent (G) |
Affects Version/s: | 2.0.0 |
Fix Version/s: | None |
Type: | New Feature Request | Priority: | Minor |
Reporter: | Roberto Rios | Assignee: | Alexei Vladishev |
Resolution: | Unresolved | Votes: | 2 |
Labels: | None | ||
Remaining Estimate: | Not Specified | ||
Time Spent: | Not Specified | ||
Original Estimate: | Not Specified | ||
Environment: |
Windows & *nix servers |
Issue Links: |
|
Description |
In my windows servers, I'm monitoring a counter called "current disk queue length". Basically, accordingly with MS: "Current Disk Queue Length is the number of requests outstanding on the disk at the time the performance data is collected." I like it because it's simple. My question is, does linux has something like this counter? I was looking into IOSTAT, but none of the counters provided seems to be equal to current disk queue lenght. The most similar is "avgqu-sz", but it's an average and the windows one is the actual value measured when the zabbix agent was contacted. Please could you register a feature request? Meanwhile you can create a simple user parameter, which would parse output of /proc or vmstat. So my request is to create an item that could monitor, either in windows and linux, the disk I/O. Something like vfs.dev.diskqueue[device]. |
Comments |
Comment by Steve mushero [ 2011 Feb 02 ] |
Yes, avgqu-sz is an average, but over 1 second; not sure you can use data more quickly than that, i.e. instant queue may not help much if it's going up and down every 10ms; I'd think an average over time would be best. For real-time, look at field 9 of /sys/block/hda/stat, etc. See: http://www.mjmwired.net/kernel/Documentation/iostats.txt |