How to rectify Linux server slowness, maybe related to swap usage?
We have a Linux server and an exclusive process run on that server. The process usually takes around 7 hours to complete but on some days we have also noticed that it took up to 10/11 hours.
We generally use vmstat command to see the memory allocations.
On days, it has ran normally, the ‘Swapped’ indicator (o/p of vmstat) showed ‘0’ whereas on days it ran slowly, the ‘Swapped’/Swpd indicator was a good number ( in M megabytes).
I would appreciate if someone can tell me if swapping is anyway related to the slowness here. If yes, then we do the system use swap memory when thousands of megabytes are still free.
How we can rectify the scenario and avoid slowness?
Any other pointers are also welcomed. Thanks in Advance!!!
Last edited by linux273; 05-21-2012 at 11:30 AM.