RE: RAC waitevent: gc freelist on one node
Date: Fri, 31 Oct 2008 08:27:38 -0700 (PDT)
Thanks for your input.
netstat -s does not show any fragmnets dropped or any erorrs. ifconfig -a does not work on HP-UX
- On Fri, 10/31/08, Bort, Guillermo <guillermo.bort_at_eds.com> wrote:
From: Bort, Guillermo <guillermo.bort_at_eds.com> Subject: RE: RAC waitevent: gc freelist on one node To: ajayoraclel_at_yahoo.com, oracle-l_at_freelists.org Date: Friday, October 31, 2008, 3:08 PM
Not sure if it fits the case, but itĘs† worth taking a look at. From the RAC FAQ:
What do I do if I see GC CR BLOCK LOST in my top 5 Timed Events in my AWR Report?
You should never see this or BLOCK RETRY events. This is most likely due to a fault in your interconnect network. Work with your system administrator or/and network administrator to find the fault. Check netstat -s
84884742 total packets received
1201 fragments dropped after timeout
3384 packet reassembles failed
You do not want to see fragments dropped or packet reassemblies failed.
eth0 Link encap:Ethernet HWaddr 00:0B:DB:4B:A2:04 inet addr:184.108.40.206 Bcast:220.127.116.11 Mask:255.255.252.0 UP BROADCAST RUNNING MULTICAST MTU:1500 Metric:1 RX packets:21721236 errors:135 dropped:0 overruns:0 frame:95 TX packets:273120 errors:0 dropped:0 overruns:0 carrier:0
You do not want to see high number of errors.
Guillermo Alan Bort
DBA / DBA Main Team
EDS, an HP company
From: oracle-l-bounce_at_freelists.org [mailto:oracle-l-bounce_at_freelists.org] On Behalf Of Ajay Thotangare
Sent: Friday, October 31, 2008 12:48 PM
Subject: RAC waitevent: gc freelist on one node
We have having a 3 node RAC and from last few days noticed performance of 1 of node has degraded. In V$session session of that node, I always see lot of sessions waiting on waitevent "gc freelist". I don't see this wait event on other 2 nodes where performance is not affected.
Not sure if this wait event is indicating something. I could not find any documentation on this wait event Can somebody please explain me this wait event or point me to right document. Does this wait event looks related to degraded performance?