Hello,
one of our slapd daemons take 99.9% of CPU time (openldap 2.3.11, bdb 4.3.29, SuSE Linux SLES-8 (i386), Kernel 2.4.21-273-smp)
With strace we detect:
sched_yield() = 0 sched_yield() = 0 ..
I found a lot about the sched_yield problem in openldap, but I'm still not sure what to do.
Any suggestions ?
Thanks !
Sepp
You should upgrade to a newer version of OpenLDAP, preferably the latest 2.3.33.
On Tue, 13 Feb 2007, Sepp wrote:
Hello,
one of our slapd daemons take 99.9% of CPU time (openldap 2.3.11, bdb 4.3.29, SuSE Linux SLES-8 (i386), Kernel 2.4.21-273-smp)
With strace we detect:
sched_yield() = 0 sched_yield() = 0 ..
I found a lot about the sched_yield problem in openldap, but I'm still not sure what to do.
Any suggestions ?
Thanks !
Sepp
Aaron Richton wrote:
You should upgrade to a newer version of OpenLDAP, preferably the latest 2.3.33.
You also should not use BDB 4.3.x, that release family was broken. Use 4.2 or 4.4 or newer. In fact the current configure scripts will refuse to work with BDB 4.3.
On Tue, 13 Feb 2007, Sepp wrote:
Hello,
one of our slapd daemons take 99.9% of CPU time (openldap 2.3.11, bdb 4.3.29, SuSE Linux SLES-8 (i386), Kernel 2.4.21-273-smp)
With strace we detect:
sched_yield() = 0 sched_yield() = 0 ..
I found a lot about the sched_yield problem in openldap, but I'm still not sure what to do.
Any suggestions ?
Thanks !
Sepp
openldap-software@openldap.org