There are several items I would really like to monitor on my CentOS servers that I cannot currently do. If I could program in C I would do this myself but since I no nothing about C I dont have any choice but to either write some scripts and add them to each server or beg for someone to add this to the agent - so here I am BEGGING!
Disk Performance
* vfs.dev read and write
- operations per second (I/Ops)
- bytes per second
* disk queue lengths
according to the manual "operations" is supossed to work but I get "Not Supported" The disk queue lengths are very important for monitoring disk subsystem perfomance amd can be found as Field #9 in /proc/diskstats
(see http://www.kernel.org/pub/linux/kern...on/iostats.txt)
CPU total utilization %
Right now I monitor the idle % and have to look at the graph "backwards" to get the total used % - it would be nice if either the item could do math (i.e. 100-idle%= used%) or if a new item was added to the agent.
Disk Performance
* vfs.dev read and write
- operations per second (I/Ops)
- bytes per second
* disk queue lengths
according to the manual "operations" is supossed to work but I get "Not Supported" The disk queue lengths are very important for monitoring disk subsystem perfomance amd can be found as Field #9 in /proc/diskstats
(see http://www.kernel.org/pub/linux/kern...on/iostats.txt)
CPU total utilization %
Right now I monitor the idle % and have to look at the graph "backwards" to get the total used % - it would be nice if either the item could do math (i.e. 100-idle%= used%) or if a new item was added to the agent.

Comment