We are wondering if we are facing performance issue in our server when running Informatica jobs. Two things to suspect:
- cache memory never comes down even when Top shows > 99% used.
- There is some contention io or network related or Cache is clogged
top - 20:58:20 up 16 days, 4:37, 16 users, load average: 7.50, 4.85, 3.82
Tasks: 386 total, 2 running, 376 sleeping, 7 stopped, 1 zombie
Cpu(s): 35.1%us, 10.1%sy, 0.0%ni, 32.5%id, 20.9%wa, 0.0%hi, 1.3%si, 0.2%st
Mem: 32877500k total, 32626692k used, 250808k free, 291804k buffers
Swap: 20971516k total, 37056k used, 20934460k free, 18906004k cached
free -g
total used free shared buffers cached
Mem: 31 31 0 0 0 18
-/+ buffers/cache: 12 18
Swap: 19 0 19
If We kill all Informatica services & jobs... the utilization in Top comes down to
top - 20:16:17 up 16 days, 3:55, 18 users, load average: 1.26, 0.94, 0.85
Tasks: 366 total, 1 running, 357 sleeping, 7 stopped, 1 zombie
Cpu(s): 8.0%us, 3.9%sy, 0.0%ni, 88.0%id, 0.1%wa, 0.0%hi, 0.0%si, 0.0%st
Mem: 32877500k total, 19378856k used, 13498644k free, 225320k buffers
Swap: 20971516k total, 37060k used, 20934456k free, 18426624k cached
A) I am not sure why the cache memory is not released ?
B) Even when utilization in Top is going to 32 GB i.e. when Informatica jobs are running, Cache is still at 18GB ... Shouldn't cache be released because Informatica jobs are hanging ?
I strongly feel this is nothing to do with Server memory because I see swap memory not come into play at all but before I go to Informatica and raise a flag... I want to make sure there is nothing from server side... what more can I do to ensure that everything on server side works fine i.e. CPU & IO or Network, as I am not able to think of any other factor affecting this from server side?