Trying to restore and got this failures any help

Well, I do not run NMM so I can't say for sure if syntax is correct or not - above will run for NMSQL, but in NMM some switches have changed.  With that in mind, first like states no full DB backup was found - can you verify with mminfo from the client that you can see full DB backup for database in question?

Hi

Yes you are right i am using NMM2.4 and here is the exact command i am using

nsrsqlrc -s "XXX" -c "XXX" -d "MSSQL$RD:AdventureWorks_2012_DEV" -C”'AdventureWorks2012_Data'='D:\SS2K12$SQL12\DBDATA\AdventureWorks2012_DEV_Data.mdf', 'AdventureWorks2012_Log'='E:\SS2K12$SQL12\DBTLOG\AdventureWorks2012_DEV_log.ldf'" -t "Mon Jun 03 12:41:52 2013" "MSSQL$RD:AdventureWorks2012"

Can you please suggest me why i am seeing the error as follwoing any specific reason why its not able to see the backups>?

51281:(pid 380):nsrsqlrc.c(7638): Could not get saveset record from mmdb; error: 0x00014054 savetime -1 not in media db

80282:(pid 380):nsrsqlrc.c(4026): Error getting backup endtime from media Database.

37740:(pid 380):Recover of ''AdventureWorks2012_Log'='E:SS2K12$SQL12DBTLOGAdventureWorks2012_DEV_log' will be skipped be cause a full backup was not found.

51281:(pid 380):nsrsqlrc.c(7638): Could not get saveset record from mmdb; error: 0x00014054 savetime -1 not in media db

80282:(pid 380):nsrsqlrc.c(4026): Error getting backup endtime from media Database.

37740:(pid 380):Recover of 'Jun' will be skipped because a full backup was not found.

51281:(pid 380):nsrsqlrc.c(7638): Could not get saveset record from mmdb; error: 0x00014054 savetime -1 not in media db

80282:(pid 380):nsrsqlrc.c(4026): Error getting backup endtime from media Database.

37740:(pid 380):Recover of '03' will be skipped because a full backup was not found.

51281:(pid 380):nsrsqlrc.c(7638): Could not get saveset record from mmdb; error: 0x00014054 savetime -1 not in media db

80282:(pid 380):nsrsqlrc.c(4026): Error getting backup endtime from media Database.

37740:(pid 380):Recover of '12:41:52' will be skipped because a full backup was not found.

51281:(pid 380):nsrsqlrc.c(7638): Could not get saveset record from mmdb; error: 0x00014054 savetime -1 not in media db

80282:(pid 380):nsrsqlrc.c(4026): Error getting backup endtime from media Database.

37740:(pid 380):Recover of '2013 MSSQL$RD:AdventureWorks2012' will be skipped because a full backup was not found.

29307:(pid 380):At least 1 valid recover target is required

Restore operation finished with error(s). Refer to the module backup log file for details.

  43709:(pid 380):Stop time: Mon Jun 03 12:44:04 2013

I did this few times with NMSQL - I assume you use NMM.  With that in respect, my suggestion might not be applicable. It seems as if your date was interpreted as DB name so I when I check the line and compare it to what I do I see following:

- for redirect I use -C switch - can't see it in your example

- path can't start with "=" - it seems as if your are missing internal DB name

- date seems to be pretty much what I use

Hi

Is there any command i can use to check the DB files

is there a way through Net-worker  command line options  to view the number of files within a backup for a database

Can you please let me know what command we need to use ?

Thank you.

You can see that with nsrinfo command.

Hi,

Sorry, was probably not related to your problem.

My issue is solved. It was due to a wrong instance name for the source db.

I wrote "somedb" instead of "MSSQL$PORTAL_PROD:somedb"

C:\>nsrsqlrc.exe -s server -c client -f -d "MSSQL$PORTAL_PROD:CopyOfsomedb" -C "'somedb'='R:\MSSQL\Data\CopyOfsomedb.mdf', 'somedb_log'='R:\MS

SQL\Data\CopyOfsomedb_log_5946.LDF'" -t "Fri May 24 18:00:00 2013" "MSSQL$PORTAL_PROD:somedb"

43708:(pid 5912):Start time: Wed Jun 05 10:23:09 2013

43621:(pid 5912):Computer Name: client User Name: XXX

                  NSR_CLIENT: client;

                  NSR_SERVER: server;

37725:(pid 5912):Recovering database 'somedb' into 'CopyOfsomedb' ...

4690:(pid 5912):RESTORE database [CopyOfsomedb] FROM virtual_device='Legato#9b49f526-6a93-4a2f-a891-be6be7f71e3b'  WITH move 'somedb' to 'R:\MSSQL\Data\CopyOfsomedb.mdf', move 'somedb_log' to 'R:\MSSQL\Data\CopyOfsomedb_log_5946.LDF', replace, norecovery

libDDBoost version: major: 2, minor: 5, patch: 2, engineering: 1, build: 360829

91650:(pid 5912): Successfully established client DDCL session for recovering save-set ID '3701380309'.

Thanks.

Hi,

I see exactly the same issue while trying to restore a SQL 2008 database. I use NMM 2.4.1.

Did you find any solution yet?

Thanks.

How did you solve the problem?

I have exactly the same problem with NMM 3.0 SP1, using nsrsqlrc.exe and -z argument.

No problem if I try to restore a database with command line without -z but if I have to use it (because for example I'm in scenario described here: https://community.emc.com/docs/DOC-37210) it returns the following error and the restore hangs:

...

51281:(pid 4936):nsr/db_apps/bsmsql/nsrsqlrc.c(7638): Could not get saveset record from mmdb; error: 0x00014054

savetime -1 not in media db

80282:(pid 4936):nsr/db_apps/bsmsql/nsrsqlrc.c(4161): Error getting backup endtime from media Database.

...

I'm sure the syntax is correct: -S norecover argument was already present in the command. I only added -z and set -t with the right nsavetime I can see with nsrinfo.exe

No problem using the same command and syntax with NMM 2.4.

Can someone help me pls?

That article applies to NMSQL, but I guess it is still valid.  How exactly did you do restore?  Because it seems like -t you supplied was not recognized.  Also you mention -z - is there a reason you are using that?  If yes, then your first restore has to be full (without -z so you use -S norecovery) and subsequent transastion logs are restored using both -S and -z with -t which corresponds to savetime of each.

Hi Hrvoje,

    thanks for your answer.

the link was only to explain the scenario: sometimes it happens someone take a Networker Full DB backup, client side, and a scheduled Transaction Logs backup intervenes in the meanwhile, or someone accindentally push a transaction log backup on disk via tsql, breaking the tlog backup chain on Networker.

In both cases, either with NMM or old NMSQL, if you want to restore to the last TLog backup, you have to go via command line, restoring every single log backup, using -z and -S norecovery, specifying the right nsavetime via -t.

Of course Full backup and TLog backup taken before the chain break can be restored without -z and with -S norecovery.

Yes, it actually seems it does not recognize nsavetime.

I take it with "C:\Program Files\EMC NetWorker\nsr\bin\nsrinfo.exe" -s -n mssql | findstr

MSSQL:/ , date=1421844425 21/01/2015 13.47.05

MSSQL:/ %/files.1421844425.1421844426, date=1421844425 21/01/2015 13.47.05

MSSQL:/ , date=1421844293 21/01/2015 13.44.53

MSSQL:/ %/files.1421844293.1421844294, date=1421844293 21/01/2015 13.44.53

MSSQL:/ , date=1421844280 21/01/2015 13.44.40

MSSQL:/ ./PRIMARY, date=1421844280 21/01/2015 13.44.40

MSSQL:/ ./PRIMARY./PenelopeGuabello2012, date=1421844280 21/01/2015 13.44.40

MSSQL:/ ./, date=1421844280 21/01/2015 13.44.40

MSSQL:/ ./PRIMARY./, date=1421844280 21/01/2015 13.44.40

MSSQL:/ %/files.1421844280.1421844281, date=1421844280 21/01/2015 13.44.40

Three backups:

- Full backup - savetime 21/01/2015 13:44:40

- Log backup - savetime 21/01/2015 13:44:53 (cannot be used in restore chain because started and ended while full was still running)

- Log backup - savetime 21/01/2015 13:47:05

Full backup restore - No problems:

"C:\Program Files\EMC NetWorker\nsr\bin\nsrsqlrc.exe" -s -c -f -d "MSSQL: " -C" ' '=' ', ' '=' '" -S norecovery -t "Wed Jan 21 13:44:40 2015" "MSSQL: "

...

37725:(pid 6524):Recovering database ' ' into ' ' ...

4690:(pid 6524):RESTORE database [ ] FROM virtual_device='Legato#b166acf1-5695-4a02-a542-2a5d94735e31'  WITH move ' ' to ' ', move ' ' to ' ', replace, norecovery

35866:(pid 6524):The restore of database ' ' completed successfully.

29310:(pid 6524):

Received  856 MB  1 item(s)  from NSR server (size information for snapshot(s) is not available).

Log backup restore - Errors and hang:

"C:\Program Files\EMC NetWorker\nsr\bin\nsrsqlrc.exe" -s -c -d "MSSQL: " -C" ' '=' ', ' '=' '" -z -S norecovery -t "Wed Jan 21 13:47:05 2015" "MSSQL: "

...

37725:(pid 4616):Recovering database ' ' into ' ' ...

51281:(pid 4616):nsr/db_apps/bsmsql/nsrsqlrc.c(7638): Could not get saveset record from mmdb; error: 0x00014054

savetime -1 not in media db

80282:(pid 4616):nsr/db_apps/bsmsql/nsrsqlrc.c(4161): Error getting backup endtime from media Database.

4690:(pid 4616):RESTORE log [ ] FROM virtual_device='Legato#94d120b0-f6b2-4981-835b-aeb7bf44d38f'  WITH norecovery

... it hangs here.

I tried with ctime

notation too (-t "1421844425"), as reported from nsrinfo, with no results.

As said it perfectly worked when tested with NMSQL and NMM2.4

Either it's a bug or they changed the way to expose nsavetime with nsrsqlrc.exe, but NMM3.0SP1 User Guide confirm the syntax.

Hope it can help.

Thank you.

I would use time from mdb instead of index as error indicates failure for timestamp using mdb.  They should be match or near, but perhaps that "near" part might be an issue.

As for other folks doing dumps, whenever there is "different initiator" for backup, SQL knows it and changes next backup to full.  This means that if your DBA did backup to disk, next scheduled transaction log backup from NetWorker will be changed to full for consistency reasons (I know this for a fact with DB dumps, not sure about TLOG, but I would assume the same).

I beg you pardon, I'm not a NSR engineer: which command should I use, client side, to request time from mdb, instead of index?

You can use mminfo -avot -q group= -r 'name,client,savetime(20),level' for example.  Or you can do the query by client:

mminfo -avot -c -r 'name,client,group,level'

Hi Hrvoje,

    thanks for your patience.

I tried with:

"C:\Program Files\EMC NetWorker\nsr\bin\mminfo.exe" -avot -s -t "last week" -q client= -r "client,name,group,savetime(20),nsavetime,sscreate(20),ssinsert(20),sscomp(20)"

Result:

   MSSQL:          DB_SQL_12h_01 21/01/2015 13.44.40 1421844280 21/01/2015 13.45.20

21/01/2015 13.45.20 21/01/2015 13.45.33

   MSSQL:          DB_SQL_12h_01 21/01/2015 13.44.53 1421844293 21/01/2015 13.45.33  21/01/2015 13.45.33 21/01/2015 13.45.33

   MSSQL:          DB_SQL_12h_01 21/01/2015 13.47.05 1421844425 21/01/2015 13.47.45 21/01/2015 13.47.45 21/01/2015 13.47.45

Times are actually different.

I tried to restore log (last row) using both server side savetime (sscreate):

"C:\Program Files\EMC NetWorker\nsr\bin\nsrsqlrc.exe" -s -c -d "MSSQL: " -C" ' '=' ', ' '=' '" -z -S norecovery -t "Wed Jan 21 13:47:45 2015" "MSSQL: "

... but it failed with the following results:

37725:(pid 4412):Recovering database ' ' into ' ' ...

51281:(pid 4412):nsr/db_apps/bsmsql/nsrsqlrc.c(7638): Could not get saveset record from mmdb; error: 0x00014054

savetime -1 not in media db

80282:(pid 4412):nsr/db_apps/bsmsql/nsrsqlrc.c(4161): Error getting backup endtime from media Database.

43336:(pid 4412): Specified backup of not found, cannot recover

35867:(pid 4412):Error processing SQL database ' '

29310:(pid 4412):

Received  0 KB  0 item(s)  from NSR server (size information for snapshot(s) is not available).

It seems it cannot find the backup at all.

Hope this can help!

Thank you

Use savetime unless savetime=sscreate.