gcorbehem at gmail.com
Tue Jul 4 17:57:29 MEST 2006
I have a big pbm with IO on my Linux server wich collects updates on
20 000 rrd files every 60 s.
A sar / iostat command shows that the server is very overloaded at
The updates are done with perl scripts.
Each rrd file has 3 RRA :
RRA:AVERAGE:0.5:1: 86400 (60 days)
RRA:AVERAGE:0.5:60: 4464 (186 days)
RRA:AVERAGE:0.5:240: 6576 (3 years)
The enormous periodic IO overload seems to come from RRD .
We tried to begin a modification on the RRA(s) to avoid those IO which seems
to come from RRD .
Those 3 RRA è 1 RRA (1 year)
3 days later the server is even more overloaded
The overload is not periodic any longer but nonstop !
Could you help me ?
Unsubscribe mailto:rrd-users-request at list.ee.ethz.ch?subject=unsubscribe
Help mailto:rrd-users-request at list.ee.ethz.ch?subject=help
More information about the rrd-users