livecache cannot be started

[former_member236442]

Hello experts,

I cannot start livecache (7.7.07.50 on Linux) due to following errors:

dbmcli on LCD>db_online

ERR

-24994,ERR_RTE: Runtime environment error

20090,Check knldiag! Kernel exited with '1' before reaching ADMIN state

The database is in state "offline", but resources were found: pid file

found but

startup message from kernel:Core dump suppressed

dbmcli on LCD>show state

OK

No root pointer in mapping table

'RTECons_MemoryDumpMapping' instance could not be created

Failed to map memory dump files

I have tried to clear pids from below files (by mv LCD LCD.old) and then start.

/sapdb/data/pid/LCD

/sapdb/data/ppid/LCD

Cleared shared memory by:

/sapdb/LCD/db/pgm/dbmshm CHECK /var/lib/sdb/dbm/ipc LCD

/sapdb/LCD/db/pgm/dbmshm DELETE /var/lib/sdb/dbm/ipc LCD

I have also tried to delete folder /sapdb/data/wrk/LCD/rtedump_dir and

checked SAPNote 1283278.

Attached is knlmsg log.

Thank you for any advice.

Ivana

knlmsg.txt.zip

[JamesZ · Accepted solution]

Hi Ivana,

Please attach diagpack.tgz after executing command below:

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

Best regards,
James

[former_member236442 · Accepted solution]

Hello James,

I cannot edit my message anymore to attach file. TGZ and neither ZIP formats are not allowed to send. Can I share it by some other way? Thanks for help.

Ivana

[JamesZ · Accepted solution]

Hi Ivana,

Change it to txt extension. Then I will download it and change to the correct format.

Best regards,
James

[former_member236442 · Accepted solution]

Hi James,

I did so. Please find attached. But I have generated diag pack after trying below solution in this discussion (by restore_param 1), not sure whether this matters.

Ivana

diagpkg.txt.zip

[JamesZ · Accepted solution]

Hi Ivana,

The error:

data volume configuration corrupted: Successor of volume 13 should be 14,_FILE=IOMan_DataFileSystem+noPIC.cpp,_LINE=2379

Please check SAP note

879058 and

1331982

.

best regards,
James

[former_member236442 · Accepted solution]

Hi James,

number of data volumes and log volumes was incosistent in parameter file and on the server. I reverted whole 'config' folder from filesystem backup made before I triggered restore_param command. Now I have the same error as reported in the beginning. I am attaching fresh diag pack (added txt suffix to be able to attach it)

dbmcli on LCD>db_state

OK

State

OFFLINE

---

dbmcli on LCD>db_online

ERR

-24994,ERR_RTE: Runtime environment error

20090,Check knldiag! Kernel exited with '1' before reaching ADMIN state

The database is in state "offline", but resources were found: pid file

found but no watchdog process

startup message from kernel:Core dump suppressed

dbmcli on LCD>db_admin

ERR

-24994,ERR_RTE: Runtime environment error

20090,Check knldiag! Kernel exited with '1' before reaching ADMIN state

The database is in state "offline", but resources were found: pid file

found but no watchdog process

startup message from kernel:Core dump suppressed

Thank you

Ivana

diagpkg.tgz.txt.zip

[JamesZ · Accepted solution]

Hi Ivana,

Can you let me know your checking result by following both notes？

Best regards,
James

[former_member236442 · Accepted solution]

Hi James,

based on SAP Notes I found out there was different number of data volumes and log volumes present on the server than in configuration. Command param_getvolsall showed 13 data volumes and 1 log volume but there were 21 data volumes and 2 log volumes on server. It means that by restore_param i "took away" data volumes from number 14. I reverted it as I described above.

I have opened OSS message to SAP and they solved the issue. The cause was that parameters MirroredLogVolumeName002 and LogVolumeName002 were identical. So I set parameter MirroredLogVolumeName002  to different path than LogVolumeName002.

Thanks a lot for help!

Ivana

[JamesZ]

Glad to solve it, parameter restore should always be used carefully

[david_liu1]

Hello,

In the attached knlmsg log:

<MSGL _NO="946096" _PROCESS="8506" _THREAD="0x213A" _TIME="2016-04-07 12:01:09.000">

  <MSG _NO="1" _TYPE="Error" _ID="20056" _COMP="RTE" _TEXT="Volume &apos;/sapdb/LCD/saplog/DISKL002&apos; already exists">

   <MSG_ARGS

When you had log full situation, you decided to add log 2, again this is not correct solution for log full situation. Adding log volume 2 most probably did not work because of log full problem.

Try following command

dbmcli -d <SID> -u control,<password> db_offline

param_restore 1

db_warm

db_online

Regards,

David

[former_member236442]

Hello David,

I tried it, but I have received another error. Please see below. However after this step I was able to start DB at least to 'admin' state which was not possible before. How can I revert param_restore 1 command?

dbmcli on LCD>param_restore 1

OK

---

dbmcli on LCD>db_warm

ERR

-24988,ERR_SQL: SQL error

-902,I/O error

3,Database state: OFFLINE

Internal errorcode, Error code 9050 "disk_not_accessible"

20017,RestartFilesystem failed with 'I/O error'

20044,data volume configuration corrupted: Successor of volume 13 should be 14

---

dbmcli on LCD>db_admin

OK

---

dbmcli on LCD>db_state

OK

State

ADMIN

---

dbmcli on LCD>db_online

ERR

-24988,ERR_SQL: SQL error

-902,I/O error

3,Database state: OFFLINE

Internal errorcode, Error code 9050 "disk_not_accessible"

20017,RestartFilesystem failed with 'I/O error'

20044,data volume configuration corrupted: Successor of volume 13 should be 14

Thanks a lot

Ivana

[former_member236442]

reverted it as described above the issue was really with log volume 2, solution is mentioned in my last reply to James