After System Copy: MaxDB won't come up

[Former Member]

Dear all,

we performed a remote homogenous System Copy MaxDB on Linux with Backup/Restore procedure. The "recover_start <medium> DATA AUTOIGNORE" succeeded with RC=0 (8-time parallel backup files).

When we try to start the Database with "db_online", the process won't terminate and the Database stays in state "ADMIN". KNLMSG doesn't get new entries, the last ones are:

Thread	0x4A81 Task	120  2015-08-19 11:10:40	dynpool	54003:  DYNP_K51_LOCK_LIST	: 90543136
Thread	0x4A81 Task	120  2015-08-19 11:10:40	dynpool	54003:  LOCK maxlocks	: 300000
Thread	0x4A81 Task	120  2015-08-19 11:10:40	dynpool	54003:  LOCK items	: 902400
Thread	0x4A81 Task	120  2015-08-19 11:10:40	dynpool	54003:  LOCK Regions	:	24
Thread	0x4A81 Task	120  2015-08-19 11:10:40	dynpool	54003:  LOCK TransRegions	:	8
Thread	0x4A81 Task	120  2015-08-19 11:10:40	dynpool	54003:  LOCK RegionGlob+Space  :   3840
Thread	0x4A81 Task	120  2015-08-19 11:10:40	dynpool	54003:  LOCK TransGlob	:	768
Thread	0x4A81 Task	120  2015-08-19 11:10:40	dynpool	54003:  LOCK SupplyItemsPerRgn :	100
Thread	0x4A81 Task	120  2015-08-19 11:10:40	dynpool	54003:  LOCK SupplySize	: 230400
Thread	0x4A81 Task	120  2015-08-19 11:10:40	dynpool	54003:  LOCK RowHash entries   : 300000
Thread	0x4A81 Task	120  2015-08-19 11:10:40	dynpool	54003:  LOCK TabHash entries   :  60000
Thread	0x4A81 Task	120  2015-08-19 11:10:40	dynpool	54003:  LOCK Row+Tab hash size :2880000
Thread	0x4A81 Task	120  2015-08-19 11:10:40	dynpool	54003:  LOCK Trans entries	:   1776
Thread	0x4A81 Task	120  2015-08-19 11:10:40	dynpool	54003:  LOCK trans_list size   : 639360
Thread	0x4A81 Task	120  2015-08-19 11:10:40	dynpool	54003:  LOCK TransPtrList	:  14208
Thread	0x4A81 Task	120  2015-08-19 11:10:40	dynpool	54003:  LOCK TransHash entries :   1776
Thread	0x4A81 Task	120  2015-08-19 11:10:40	dynpool	54003:  LOCK TransHash size	:  14208
Thread	0x4A81 Task	120  2015-08-19 11:10:40	dynpool	54003:  LOCK SupplyPoolSegments:   9000
Thread	0x4A81 Task	120  2015-08-19 11:10:40	dynpool	54003:  LOCK SupplyPoolSize	: 86616000

"db_cons show task" shows one interesting task

T120   8  19073 User

19035 IO Wait (R)

10

45

54(r)

--> what is he waiting for?

"param_checkall" says OK.

Any helpful hints will be greatly appreciated. The target Database Release is the same as in Source: 7.8.02.035 (we took the 7.8.02.031 installation package and patched it).

Kind regards,

Dirk

[JamesZ · Accepted solution]

Hi Dirk,

Please attach diagpack.tgz file after execute below os command:

dbmcli -d <dbsid> -u control,<password> diag_pack

You can find diagpack.tgz from maxdb run directory afterwards.

Best regards,

James

[Former Member · Accepted solution]

Hi James,

thanks for your effort, points will be rewarded!

We found that the log partitions between source and target system were different: The source system has 4 log volumes with 2 GB each in 1 partition. The target system also has 4 log volumes with 2 GB each but only log volume #1 (DISKL001) belonging to log partition #1 (other log volumes were sort of unattached). We fixed this by putting log volumes #2 to #4 into log partition #1 via param_directput.

After a new "db_activate recover <medium> data autoignore" the database finally came up!