This post is more than 5 years old
6 Posts
0
10043
SQL server backup fails with return code 11
Hello all,
I have a problem with a new SQL server. We have defined it as a client using the New Client Wizard. We have given the correct authorization in Configuration -> User Groups and installed the client and networker module for Microsoft on the client. When executing the backup from the Networker server we get the following error in the Networker Console:
Exited with exit code: 11, completion severity: INFORMATION(10), completion status: unexpectedly exited(2)
The messages from the command itself does not help me much:
# savegrp -vvvvvv Ekstra
90528:savegrp: sqlserver.domain:MSSQL: level=incr
7236:savegrp: Group will not limit job parallelism
83643:savegrp: sqlserver.domain:savefs started
savefs -s nwserver.domain -c sqlserver.domain -g Ekstra -p -l full -R -v -F MSSQL:
3:savegrp: Group Ekstra waiting for 1 jobs (0 awaiting restart) to complete.
92800:savegrp: Unable to determine status for job 783688, sqlserver.domain:savefs: Log file '/nsr/logs/sg/Ekstra/783688' not found.
84078:savegrp: command 'savefs -s nwserver.domain -c sqlserver.domain -g Ekstra -p -l full -R -v -F MSSQL:' for client sqlserver.domain exited with return code 11
90491:savegrp: sqlserver.domain:savefs unexpectedly exited.
97274:savegrp: Stopped processing of the save set sqlserver.domain:MSSQL: because the savefs job job terminated with an error.
77562:savegrp: job (783689) host: sqlserver.domain savepoint: MSSQL: had ERROR indication(s) at completion
90491:savegrp: sqlserver.domain:MSSQL: abandoned.
83643:savegrp: sqlserver.domain:index started
save -s nwserver.domain -S -g Ekstra -LL -f - -m nwserver.domain -V -t 1413880440 -l 9 -W 78 -N index:1bfd43e3-00000004-53a952d1-5446091c-e05f9dd8-30806a67 /nsr/index/sqlserver.domain
<
<< / >>
forget
ignore
+uasm: db6
+uasm: db6 ????0000
+mtimeasm: *.rec
+skip: * .?*
90491:savegrp: sqlserver.domain:index succeeded.
83647:savegrp: Ekstra sqlserver.domain:index See the file /nsr/logs/sg/Ekstra/783690 for command output
7241:savegrp: nsrim run recently, skipping
The logfile /nsr/logs/sg/Ekstra/783690 only contains the index:
libDDBoost version: major: 2, minor: 6, patch: 1, engineering: 1, build: 394260
86704:save: Successfully established DDCL session for save-set ID '2437391804' (nwserver.domain:index:1bfd43e3-00000004-53a952d1-5446091c-e05f9dd8-30806a67).
75146:save: Saving files modified since Tue Oct 21 10:34:00 2014
/nsr/index/sqlserver.domain/db6/54470000/544786ad.rec
/nsr/index/sqlserver.domain/db6/54470000/544786fb.rec
/nsr/index/sqlserver.domain/db6/54470000/54477755.rec
/nsr/index/sqlserver.domain/db6/54470000/54477758.rec
/nsr/index/sqlserver.domain/db6/54470000/544776e4.rec
/nsr/index/sqlserver.domain/db6/54470000/5447768a.rec
/nsr/index/sqlserver.domain/db6/54470000/54477752.rec
/nsr/index/sqlserver.domain/db6/54470000/544795c2.rec
/nsr/index/sqlserver.domain/db6/54470000/54479619.rec
/nsr/index/sqlserver.domain/db6/54470000/5447774d.rec
/nsr/index/sqlserver.domain/db6/54470000/
/nsr/index/sqlserver.domain/db6/54460000/5446f150.rec
/nsr/index/sqlserver.domain/db6/54460000/5446e674.rec
/nsr/index/sqlserver.domain/db6/54460000/5446eb9a.rec
/nsr/index/sqlserver.domain/db6/54460000/5446e657.rec
/nsr/index/sqlserver.domain/db6/54460000/5446ebd2.rec
/nsr/index/sqlserver.domain/db6/54460000/5446e924.rec
/nsr/index/sqlserver.domain/db6/54460000/
/nsr/index/sqlserver.domain/db6/
/nsr/index/sqlserver.domain/
nwserver.domain: index:sqlserver.domain level=9, 204 KB 00:00:03 20 files
completed savetime=1413981627
According to the sql admin there is nothing in the networker log on the sqlserver.
Any suggestion, anyone?
crazyrov
4 Operator
4 Operator
•
1.3K Posts
0
October 23rd, 2014 04:00
The SYSTEM account by default is the account with which the nsrexecd service runs and is the logon used to request access to the SQL server. In-case the SYSTEM account does not have sysadmin privileges on your SQL application it is unable to log in and accessthe databases to initiated the backups. Thus it fails.
crazyrov
4 Operator
4 Operator
•
1.3K Posts
0
October 22nd, 2014 06:00
Hi,
The probe job is failing. Which means your this is a communication issue or a license issue. Check if tee name of the backup server is mentioned in the server file located at C:\Program Files\EMC NetWorker\nsr\res\servers. if not make an entry and restart the nsrexecd servcies for the changes to take affect. Test a filessytem backup.
I would also check the xbsa.messages file to see if there is anything of importance that is logged in it.
epimetheus1
6 Posts
0
October 22nd, 2014 10:00
Hi,
The filesystem backup works OK. Only the SQL backup fails.
The event log on the sql server has an Application Error:
Faulting application name: savefs.exe, version 8.2.0.1, time …
Faulting module name: liblocal.dll, version 8.2.0.1, time stamp …
Exception code: 0xc0000005
...
I have also checked the xbsa.messages as you suggested, and there are some of these messages there:
_nwbsa_is_retryable_error: received a network error (Severity 0 Number -7): no matching devices for save of client `sqlserver'; check storage nodes, devices or pools
Does this help?
ble1
14.3K Posts
0
October 22nd, 2014 11:00
a) what is the backup command you use for SQL backup?
b) did you try the same from client?
epimetheus1
6 Posts
0
October 22nd, 2014 11:00
I have created a test group for the SQL server, and I started the backup with the savegrp-command. As the job fails already on the probe, the actual backup has not been started. probe
I have tried the probe on the client with the command: "savefs -s nwserver -c sqlserver -g Ekstra -p -l full -R -v -F MSSQL:". The command just returns to the prompt and does not seem to do anything.
ble1
14.3K Posts
0
October 22nd, 2014 13:00
You didn't answer a) and you didn't understand b) - I wish to know what do you get if you start SQL backup from client and not probe.
I also see that you run 8.2.0.1 - why oh why Well, in such case I assume you also run NMM 8.2? I would suggest to try something more mature and market present 8.1.x with NMM 3.1SP1. 8.2 is only in its GA release now for a month or so (first patch will be released this week) and I would use it only if I had to. Do you have to?
epimetheus1
6 Posts
0
October 22nd, 2014 22:00
SQL backup from the client work OK if started from "Networker User for SQL server".
We have been recommended by an EMC partner to use the last available module and client. I will check with them if we can use the older versions, and if so I will try to see if they work better.
PS.
Sorry if I seem stupid to you. When you asked me what command I used to take backup I thought you ment what command I used to take the backup. Sorry about that. I should have understood that you ment something else.
crazyrov
4 Operator
4 Operator
•
1.3K Posts
0
October 23rd, 2014 00:00
The backups work from the client and not from the server end. Well something to do with the credentials? Can you try putting in the creds that you used run the client initiated backup into the client configurations.
guenterH1
117 Posts
1
October 23rd, 2014 00:00
As I understand from what has been told here so far is you are doing a "classic" SQL-VDI backup and not a VSS snapshot backup of SQL Server.
In such case it is common to have a "remote user" and password set in the SQL client backup configuration.
Mostly due to the fact that in newer versions of SQL Server the entry of local SYSTEM has been removed by Microsoft from the authorized users list of SQL Server.
Such entry of remote user may be authorized to access SQL-DB for backup but it may be prevented by Windows UAC from running the savefs command.
Do you have a remote user setup here?
If so try (just for a test) to add the local SYSTEM as a full access account to the SQL Server logons and remove all related entries in your NetWorker client configuration.
The other test would be disabling Windows UAC on the SQL server host completely.
epimetheus1
6 Posts
0
October 23rd, 2014 04:00
Thank you, that explains it. With correct priveleges for SYSTEM the backup now work fine.
epimetheus1
6 Posts
0
October 23rd, 2014 04:00
It seems that you are on to something here. When we did the first test (add the local SYSTEM as a full access account) then the backup worked!
My SQL admin has created a domain user with extended authorizations and executes the networker services with this user, and now the SQL backup works fine.
Does this seem right to you?
BTW: UAC is off.
guenterH1
117 Posts
0
October 23rd, 2014 05:00
My experience about UAC is: turned off doesn't always mean really "off".
Thai also depends very much on AD global group policy settings.
In certain situation it might be the case that even UAC is turned off you get a request (pop-up window) to authorize the elevation for a command. This even your account has all administrative rights and groups assigned you can think of.
If this is the case in any context on this server the savefs command will fail to run.