Dirty Buffer Pool page

[JonRoenick]

If you read this forum often you'll see i've been struggling to get this TSM healthy lately. Every week brings something new. Today started with an expiration running for only 2 hours, and stopping successfully after only 200k objects (it should be over 5.2million). So started expiration again and it started to run better, then i noticed the recovery log jumped from 0% to 30% in a matter of minutes. Here are some outputs:

show logpin:
Dirty page Lsn=4805528.166.3788, Last DB backup Lsn=4808197.152.2662,
Transaction table Lsn=0.0.0, Running DB backup Lsn=0.0.0, Log truncation
Lsn=4805528.166.3788
Lsn=4805528.166.3788, Owner=DB, Length=518
Type=Update, Flags=C2, Action=ExtDelete, Page=4852651, Tsn=0:3905635955,
PrevLsn=4805528.166.3456, UndoNextLsn=0.0.0, UpdtLsn=4805408.191.3451 ===>
ObjName=Backup.Objects, Index=0, RootAddr=78, PartKeyLen=9, NonPartKeyLen=190,
DataLen=219
The recovery log is pinned by a dirty page in the data base buffer pool. Check
the buffer pool statistics. If the associated transaction is still active then
more information will be displayed about that transaction.
Database buffer pool global variables:
CkptId=115385, NumClean=86955, MinClean=196512, NumTempClean=196608,
MinTempClean=98256,
BufPoolSize=196608, BufDescCount=216268, BufDescMaxFree=216268,
DpTableSize=196608, DpCount=109653, DpDirty=109653, DpCkptId=114492,
DpCursor=57002,
NumEmergency=0 CumEmergency=0, MaxEmergency=0.
BuffersXlatched=0, xLatchesStopped=False, FullFlushWaiting=False.


q db f=d
Available Space (MB): 100,000
Assigned Capacity (MB): 100,000
Maximum Extension (MB): 0
Maximum Reduction (MB): 9,804
Page Size (bytes): 4,096
Total Usable Pages: 25,600,000
Used Pages: 10,059,545
Pct Util: 39.3
Max. Pct Util: 39.5
Physical Volumes: 40
Buffer Pool Pages: 196,608
Total Buffer Requests: -1,449,133,283
Cache Hit Pct.: 99.47
Cache Wait Pct.: 0.00
Backup in Progress?: No
Type of Backup In Progress:
Incrementals Since Last Full: 0
Changed Since Last Backup (MB): 1,149.65
Percentage Changed: 2.93
Last Complete Backup Date/Time: 05/19/2010 05:36:42
Estimate of Recoverable Space (MB):
Last Estimate of Recoverable Space (MB):


---------------

I'm not sure what i'm supposed to be looking at. I'm just assuming that a negative 1.4 billion buffer requests is a bad thing, but i don't know what it means or how to fix it. Meanwhile i have canceled expiration and no other processes or sessions are running, yet the log continues to go up.

 

[GregE]

Have you restarted TSM recently?

Here's my show logpinned, for reference (my db is in roll forward mode)

Dirty page Lsn=4622893.174.786, Last DB backup Lsn=4617535.106.464, Transaction table Lsn=4622892.222.3196, Running DB backup Lsn=0.0.0, Log truncation Lsn=4617535.106.464

Lsn=4617535.106.464, Owner=DB, Length=203
Type=Update, Flags=C2, Action=ExtInsert, Page=146, Tsn=0:1574014063, PrevLsn=0.0.0, UndoNextLsn=0.0.0, UpdtLsn=4617535.105.3775 ===> ObjName=Activity.Log, Index=2, RootAddr=56,
PartKeyLen=0, NonPartKeyLen=8, DataLen=95
The recovery log is pinned by the last data base backup. Space will not be freed until data base backup is run again.

 

[JonRoenick]

10:33:33 TSM : show logpin cancel
Dirty page Lsn=4805528.166.3788, Last DB backup Lsn=4808200.219.355,
Transaction table Lsn=0.0.0, Running DB backup Lsn=0.0.0, Log truncation
Lsn=4805528.166.3788
Lsn=4805528.166.3788, Owner=DB, Length=518
Type=Update, Flags=C2, Action=ExtDelete, Page=4852651, Tsn=0:3905635955,
PrevLsn=4805528.166.3456, UndoNextLsn=0.0.0, UpdtLsn=4805408.191.3451 ===>
ObjName=Backup.Objects, Index=0, RootAddr=78, PartKeyLen=9, NonPartKeyLen=190,
DataLen=219
The recovery log is pinned by a dirty page in the data base buffer pool. Check
the buffer pool statistics. If the associated transaction is still active then
more information will be displayed about that transaction.
Database buffer pool global variables:
CkptId=115386, NumClean=119463, MinClean=196512, NumTempClean=196608,
MinTempClean=98256,
BufPoolSize=196608, BufDescCount=216268, BufDescMaxFree=216268,
DpTableSize=196608, DpCount=77145, DpDirty=77145, DpCkptId=114492,
DpCursor=89514,
NumEmergency=0 CumEmergency=0, MaxEmergency=0.
BuffersXlatched=0, xLatchesStopped=False, FullFlushWaiting=False.

 

[JonRoenick]

Have you restarted TSM recently?

May 9th was last time it was restarted.

 

[JonRoenick]

I use Normal mode, not rollforward

 

[moon-buddy]

I use Normal mode, not rollforward

Any particular reason why normal instead of rollforward? Normal mode does not allow for db backup trigger (in reference to your earlier related posting).

 

[JonRoenick]

That is how the consultants set it up. I read the pros and cons from the link in post 5, but it didn't seem to have a real benefit for me that i could see. Plus, i'm out of tapes right now so i have no way to backup the database.

I've been fighting with this expiration for quite some time so reclamation doesn't get to run fully. I swear there is something wrong b/c it keeps getting hung up with dirty buffer pool pages.

 

[moon-buddy]

That is how the consultants set it up. I read the pros and cons from the link in post 5, but it didn't seem to have a real benefit for me that i could see. Plus, i'm out of tapes right now so i have no way to backup the database.

I've been fighting with this expiration for quite some time so reclamation doesn't get to run fully. I swear there is something wrong b/c it keeps getting hung up with dirty buffer pool pages.

Try moving to rollforward for now.