DB2 - Problem description
| Problem IC92805 | Status: Closed |
FIX EMP CHAIN WALKING CODE IN INCREMENTAL BACKUP CODE PATH | |
| product: | |
DB2 FOR LUW / DB2FORLUW / A10 - DB2 | |
| Problem description: | |
Incremental backup with tablespaces that have a very large
number of non-headered pages (ie. type lob, clob, etc), it can
fail with a -902 error. Stack traces will indicate
sqlbMapIndirectEMPs, and diagprobes regarding bad page reads
might occur...
Example Stack
0] 0x00002AC7188E9842 pdLog + 0x230
[1] 0x00002AC719100E68 _Z19sqlbGetPageFromDiskP11SQLB_FIX_CBi
+ 0x1754
[2] 0x00002AC718E39940 _Z7sqlbfixP11SQLB_FIX_CB + 0x4B70
[3] 0x00002AC7182E0C0E _Z11sqlbFixPageP11SQLB_FIX_CBiji +
0x33A
[4] 0x00002AC718EDB11D
_Z19sqlbMapIndirectEMPsP12SQLB_GLOBALSP11SQLB_FIX_CBjPcjP16SQLB_
OBJECT_DESCP12SQLB_POOL_CB + 0x1F1
[5] 0x00002AC718EDBCE2
_Z24sqlbMapLongObjectExtentsP12SQLB_GLOBALStPPcPj + 0x884
[6] 0x00002AC71A47F88C _Z12sqlubReadDMSP14SQLU_BUFMAN_CB +
0x89C
[7] 0x00002AC71A47C5B4 _Z18sqlubBMProcessDataP14SQLU_BUFMAN_CB
+ 0x9B8
[8] 0x00002AC71A47A750 _Z8sqlubbufP8sqeAgent + 0xA90
[9] 0x00002AC7198E55D6
_Z26sqleSubCoordProcessRequestP8sqeAgent + 0x1AC
[10] 0x00002AC71991FA79 _ZN8sqeAgent6RunEDUEv + 0x3EB
[11] 0x00002AC71ECAC58C _ZN9sqzEDUObj9EDUDriverEv + 0x17C
[12] 0x00002AC71ECAC1B4 _Z10sqlzRunEDUPcj + 0x26
[13] 0x00002AC71C4A6DC3 sqloEDUEntry + 0x971
[14] 0x00002AC7167692A3 /lib64/libpthread.so.0 + 0x62A3
[15] 0x00002AC7252116DD __clone + 0x6D
Example DiagProbe
2013-05-14-19.42.48.963713-240 I187893E6851 LEVEL:
Warning
PID : 3915 TID : 47035097016640 KTID : 6420
PROC : db2sysc
INSTANCE: shawnrr NODE : 000 DB : EMP1
APPHDL : 0-76 APPID:
*LOCAL.shawnrr.130514234237
AUTHID : SHAWNRR HOSTNAME: hotellnx99
EDUID : 299 EDUNAME: db2bm.277.1 (EMP1)
FUNCTION: DB2 UDB, buffer pool services, sqlbGetPageFromDisk,
probe:932
MESSAGE : ZRC=0x87020036=-2029912010=SQLB_BADHDR "Bad Page
Header"
DIA8547C An error occurred in a database page header.
DATA #1 : String, 39 bytes
Page from disk doesn't match obj or BPD
DATA #2 : Page header, PD_TYPE_SQLB_PAGE_HEAD, 48 bytes
...
The number of pages required to encounter this problem is based
on both the page size and the extent size, with the issue being
more likely to be seen if both page size and extent size are
small. The issue requires there to be Level 2 EMP pages created
to map the pages in the tablespace which is a factor of the
afforementioned conditions.
As an example for 4K pages and an extent size of 2 (the smallest
possible combo), one requires a minimum of 3939970 pages to have
the possibility of seeing this problem.
For an 8K page, the number of pages increases to 16297114. | |
| Problem Summary: | |
**************************************************************** * USERS AFFECTED: * * Incremental backup users with lob objects and very large * * numbers of used pages in tablespace * **************************************************************** * PROBLEM DESCRIPTION: * * See Error Description * **************************************************************** * RECOMMENDATION: * * Upgrade product to latest fixpac. * **************************************************************** | |
| Local Fix: | |
Move to latest fixpack. | |
| available fix packs: | |
DB2 Version 10.1 Fix Pack 3 for Linux, UNIX, and Windows | |
| Solution | |
Fixed in this fixpac of db2. New behavior is that incremental backups will succeed | |
| Workaround | |
not known / see Local fix | |
| Timestamps | |
Date - problem reported : Date - problem closed : Date - last modified : | 04.06.2013 29.08.2014 29.08.2014 |
| Problem solved at the following versions (IBM BugInfos) | |
| Problem solved according to the fixlist(s) of the following version(s) | |
| 10.1.0.3 |
|
| 10.1.0.3 |
|