on 11-15-2011 11:01 AM
Hello.
We are experiencing a huge amount of collisions in region KMStripe for our Livecache production instance (APO).
I cannot explain this or find that it has any impact on the system overall.
Is there any reason to be alerted by these numbers and what can we do ?
Livecache version: 7.7.04 BUILD 036-123-213-661
ID NAME Col.Rate Collisions Waits Accesses
-
223 KMStripe 85,21 4.953.632 50.131 5.813.598
Hi,
Could you please check the following
1) Number of lock entries configured
2) Lock pool size
3) Data cache configuration.
From your scenario it looks like bottleneck in lock management area.
Regards,
Deepak Kori
You must be a registered user to add a comment. If you've already registered, sign in. Otherwise, register and sign in.
i cannot see any problems in the "Lock activity" area.
The statistics look fine to me ?
What i can see is waits and Log I/O Q.overflow under "Logging activity".
Lock Activity
Available Entries 451.200 Row Locks 28.615.703
Maximum set 37.950 Table Locks 112.448.392
Average set 26
Lock Owner 0 Collisions 223
Lock Requester 0 Escalations 0
Logging Activity
Log Pages Written 11.075.372 Group commits 0
Waiting for Log Writer 8.435.416 Log I/O Queue Overflow 138.220
Scan and Sort Activity
Table Scans 36.647.284 Cache Sorts 0
Index Scans 0 Row Sorts 0
User | Count |
---|---|
85 | |
10 | |
10 | |
10 | |
7 | |
6 | |
6 | |
5 | |
4 | |
4 |
You must be a registered user to add a comment. If you've already registered, sign in. Otherwise, register and sign in.