Quantcast
Channel: Data Protector Practitioners Forum topics
Viewing all 3189 articles
Browse latest View live
↧

(DP) Support Tip: Configuring a backup to use a specific LAN.

February 3, 2016, 5:49 am
$
0
0

   Assume you have a Media Agent on client name media_agent.com, the official hostname, and there is a second LAN on the same host with the address media_agent_bkup.com which you which to use to perform backups.

   Ensure the Data Protector client is installed on media_agent.com and media_agent_bkup.com is imported as a virtual host. Configure a Media Agent for both client names media_agent.com and media_agent_bkup.com. When configuring a backup, choose the Media Agent configured on the client which corresponds to the LAN you which to use for backing up the data. The Session Manager will tell the Disk Agent to connect to the Media Agent via the backup LAN. It is possible to configure a backup so that multiple objects on a backup host can be backed up across both LANs concurrently by configuring both Media Agents in the same backup.

   It is important to note the client name of the Disk Agent does not determine which LAN is used for data backup. The LAN used for data backup is based on the client name configured for the Media Agent. Also, when a session is run, the Media Agent is reported as starting on the official hostname even if connections are via a different LAN.

 

↧
Search

(DP) Support Tip: Increasing backup performance of UNIX client.

February 3, 2016, 5:59 am
$
0
0

   If a UNIX client's file system is on an AdvancedJFS, performance of a serial read of a file may increase if the file system buffer is bypassed. To enable a backup to do this, set the following in the .omnirc file on the backup host.

OB2VXDIRECT=1

By default this is turned on for Solaris file system backups but needs to be explicitly set in the .omnirc file for other UNIX hosts. This will also enable the feature for SAP R/3 integration backups.

 

↧
↧

Trouble migrating DP 6.11 from Win2003 32-bit to Win2006R2 64-bit

February 3, 2016, 11:09 am
$
0
0

SUMMARY:
When I do the actual
   winomnimigrate.pl -configure -keep_dcdirs
it seems to run fine, no errors,
but nothing is migrated.
It also runs in just 30 seconds.

I've read several online posts, but nothing specifically helpful.
I also have a support call in to HP


DETAILS:

Current DPcell2003 installation
on the D: drive, so everything is in:

D:\Program Files\OmniBack\Config
& D:\Program Files\OmniBack\db40
.

The steps I did were:

1) Install DP6.11 on DPcell2008

code in D:\Program Files :
  D:\Program Files\OmniBack

data in D:\ProgramData\OmniBack :
  D:\ProgramData\OmniBack\Config
  D:\ProgramData\OmniBack\db40
.

2) on OLD DPcell2003
  omnisv stop

3) copy
   D:\Program Files\OmniBack\Config\server
& D:\Program Files\OmniBack\bd40
to temp location on the DPcell2008

4) on OLD 2003 DPcellserver
  omnisv start

5) on NEW DPcell2008
copy the temp IDB/config to *OLD* location on DPcell2008 :
   D:\Program Files\OmniBack\Config\server
& D:\Program Files\OmniBack\bd40
(NOTE "Program Files" *not* "ProgramData"

6) on NEW DPcell2008, run the migrate
(on DPCELL2008: D:\)
> cd "Program Files\OmniBack"\bin
> perl winomnimigrate.pl -configure -keep_dcdirs

NOTE: winomnimigrate.pl started with options -configure -keep_dcdirs at Wed Feb 3 11:35:13 2016
Setting RDSDLL variable...

NOTE: Converting 32-bit IDB catalog...
HP Data Protector services successfully stopped.
HP Data Protector services successfully started.

NOTE: Changing IDB owner...
Successfully changed cell name.
Successfully remaped Detailed Catalog directories.

NOTE: Configuring users.
NOTE: Updating local cell_info with new CM information.
HP Data Protector services successfully stopped.
HP Data Protector services successfully started.

NOTE: Configuring clients.
Successfully updated information for client system "dpcell2008.xxx.yyy"
and changed client's cell.


Runs in ~ 30 seconds,
no errors

BUT
Notice that it only found client "dpcell2008.xxx.yyy".

Nothing else is in the new Cell except the dpcell2008,
and the D:\Program Files\OmniBack\bd40 is very tiny
so the IDB did not get migrated.

 

bv

↧

Cannot BACKUP or RESTORE a database snapshot

February 3, 2016, 1:25 pm
$
0
0

<I have problem with sql integration, at backup time show me the following error:

 

[Critical] From: OB2BAR_SQLBAR@sqlbicl.ccr.co.cr "(DEFAULT)"  Time: 2/1/2016 5:04:34 AM

      Error has occurred while executing a SQL statement.

      Error message: 'SQLSTATE:[42000] CODE:(3002) MESSAGE:[Microsoft][ODBC SQL Server Driver][SQL Server]Cannot BACKUP or RESTORE a database snapshot.

SQLSTATE:[42000] CODE:(3013) MESSAGE:[Microsoft][ODBC SQL Server Driver][SQL Server]BACKUP DATABASE is terminating abnormally

↧

(DP) Support Tip - Some advice using WORM Media with Data Protector.

February 4, 2016, 1:48 am
$
0
0

When using Write Once Read Many (WORM) media with Data Protector (DP) please be aware of the following

1. Make sure the Operating System of the DP Media Agent server and the WORM media type used are listed as supported in the DP Device Support Matrix for the DP version you will use them with.

2. For some DP versions, particularly DP 9.0 at the time of writing, an omnirc setting is required on the DP Media Agent systems when using WORM media... OB2WORMDETECT=1  (default is 0).

By default (for windows and unix), the omnirc (.omnirc on unix) file can be found as follows...
\ProgramData\OmniBack\omnirc - Windows 2008 and later,
/opt/omni/.omnirc - unix systems

↧
↧

Restore Session fails after second tape in DP9.00

February 4, 2016, 4:49 am
$
0
0

Hi folks,

In the last days I try to perform a 7.5Tb restore of file system.
The restore run perfectly until the drive asks the next tape, after that I got the following error:

[Normal] From: RMA@xxx-xxx-xxx.domain.local "DRIVE 3 - MSL4048 2"  Time: 04/02/2016 09:54:08
      Ejecting medium '18'.

[Normal] From: RMA@xxx-xxx-xxx.domain.local "DRIVE 3 - MSL4048 2"  Time: 04/02/2016 09:54:08
      => UMA@xxx-xxx-xxx.domain.local@Changer3:0:6:1
      Unloading medium to slot 18 from device Tape6:0:10:0C

[Normal] From: RMA@xxx-xxx-xxx.domain.local "DRIVE 3 - MSL4048 2"  Time: 04/02/2016 09:54:40
      => UMA@xxx-xxx-xxx.domain.local@Changer3:0:6:1
      Loading medium from slot 2 to device Tape6:0:10:0C

[Critical] From: VRDA@xxx-xxx-xxx.domain.local "E: [FILE SERVER1_EVA4100]_0005"  Time: 04/02/2016 09:55:40
      Missing of data block detected, expecting block 482328 !

[Critical] From: VRDA@xxx-xxx-xxx.domain.local "E: [FILE SERVER1_EVA4100]_0005"  Time: 04/02/2016 09:55:40
      Error reading next backup record !

[Warning] From: VRDA@xxx-xxx-xxx.domain.local "E: [FILE SERVER1_EVA4100]_0005"  Time: 04/02/2016 09:55:40
      Restore terminated with errors !

[Normal] From: VRDA@xxx-xxx-xxx.domain.local "E: [FILE SERVER1_EVA4100]_0005"  Time: 04/02/2016 09:55:40
      ABORTED Disk Agent for bhz-fls-app2.mrvbh.local:E:

[Normal] From: RMA@xxx-xxx-xxx.domain.local "DRIVE 3 - MSL4048 2"  Time: 04/02/2016 09:56:17
      Ejecting medium '2'.

 

I've done several tests, when the sequential tape is requested, the error occurs.
The backup is always completed successfully, the problem is in restoration with 2 or more tapes.

 

Thanks.

 

↧

I can't restore dp6 backup with dp7

February 4, 2016, 2:39 pm
$
0
0

hello everybody,

I just entry to work in a new entreprise, it work with dp7, and today I must restore a old backup, but it is make the backup with dp6, I put cartridge inside the tape library robot, and dp7 see them but I can't import catalog and see objets inside them, there are a question mark in the cartridge and only I can format them, there are any way to restore dp6 backup from dp7?

thanks, and sorry for my english.

Michael A.

↧

(DP) Support Tip- Backup or Restore SQL database snapshot

February 5, 2016, 1:12 am
$
0
0

Hello

If you have any SQL snapshot database created manually from SQL query is not possible to do backup or restore with Data Protector.

Error message: 'SQLSTATE:[42000] CODE:(3002) MESSAGE:[Microsoft][ODBC SQL Server Driver][SQL Server]Cannot BACKUP or RESTORE a database snapshot

As workaorund you must exclude this database\snapshot from backup or remove it.

Best Regards

↧
Search

Restore problem HP DP 9.0.5

February 5, 2016, 2:15 am
$
0
0

I have a problem with resoring Filetype backup from StorOnce softwre deduplication volume mounted on media server.

While restoring with wizzard on source tab , displays warning Internal error: DbaXXXX functions. (can't read contents -> gray out).

Linux Cell_manager ( same system MA i StorOnceSoftware dedup) 9.0.5 version.

 

Thanks for any help or direction where to look for workaround.

 

↧
↧

Drive-based encryption problem after upgrade to HP DP 9.0.5

February 5, 2016, 3:41 am
$
0
0

Linux based CM,IS, MA  after upgrade from 9.0.4 to 9.0.5 drive-base encryption no longer work. With backups and with importing tape already encrypted (after importing key to IDB ).

 

 

↧

(DP) Support Tip: PANIC: could not open control file "Global/pg_control" Permission Denied.

February 5, 2016, 7:15 am
$
0
0

in instances where the IDB service fails to start for example after an upgrade and a manual atempt to start the service

ex;

D:\Omniback\idb\bin>pg_ctl.exe start –D “d:\Omniback\server\db80\pg

reports the following:

PANIC: could not open control file "Global/pg_control" Permission Denied.

check the PG_logs (ProgramData/omniback/Server/DB80/PG/PG_log) (/Var/opt/omni/server/db80/pg/pg_log)

should you see the following entry or similiar.


2016-01-26 14:35:37 CET LOG: database system was interrupted; last known up at 2015-11-09 18:55:13 CET
2016-01-26 14:35:37 CET LOG: could not open file "pg_xlog/000000010000000400000041" (log file 4, segment 65): No such file or directory
2016-01-26 14:35:37 CET LOG: invalid checkpoint record
2016-01-26 14:35:37 CET FATAL: could not locate required checkpoint record
2016-01-26 14:35:37 CET HINT: If you are not restoring from a backup, try removing the file "D:/Omniback/server/db80/pg/backup_label".
2016-01-26 14:35:37 CET LOG: startup process (PID 7996) exited with exit code 1
2016-01-26 14:35:37 CET LOG: aborting startup due to startup process failure

rin this scenario emoving the file "backup_label" resolved the issue and the IDB service was able to start. 

Technical Notes:

By default, DataProtector has continous archiving enabled in postgres.conf file:
 
archive_command = 'copy "%p" "D:/ProgramData/Omniback/server/db80/pg/pg_xlog_archive/%f"'
archive_mode = on
 
Which means that write ahead log (WAL) in the pg_xlog folder is stored to that previously mentioned folder.
 
WAL files contain every change made to the database and should the Server crash then postgres uses the WAL logs to perform a recovery.
 
As seen above in the postgres logs:
 
database system was interrupted; last known up at 2015-11-09 18:55:13 CET
 
An unexpected event did occur and postgres was trying to replay the WAL logs. However it could not find the referenced xlog files:
 
 could not open file "pg_xlog/000000010000000400000041" (log file 4, segment 65): No such file or directory
 
and also the backup history file was still present (pg_label file).
 
The reasons why the pg_label file remained on the system could be:that the server was stopped manually with pg_ctl stop, while idb backup is running but in this instance is unknown.
 
 
 

 

 

 

↧

DP 7 corrupt IDB. writedb fails

February 5, 2016, 7:35 am
$
0
0

Log for omnidbcheck has:

Checking media management database...
  * Possible bad next_slot value (2232    ) in file # 0
Key file devices.key: total 1 nodes
Key file stores.key: total 1 nodes
Key file pools.key: total 1 nodes
Key file media.key: total 24 nodes
Key file barcode.key: total 30 nodes
Key file desc.key: total 64 nodes
Data file devices.dat: total 1915 slots for records

Data file devices.dat: problems at record 145:
  * record is not deleted, but is on delete chain
  * record is deleted, but is not on delete chain
  * record is deleted, but is not on delete chain
  * record's dba is inconsistent: [0:0]
  * key field DEVNAME(32768) error: has a missing key
  * record's dba is inconsistent: [0:0]
  * key field DEVNAME(32768) error: has a missing key
  * record's dba is inconsistent: [0:0]
  * key field DEVNAME(32768) error: has a missing key

There are many of those lines repeating and then...

  * record's dba is inconsistent: [0:0]
  * set STORESTODEVICES(0) error:
    owner-pointer is null, but next and previous are not
  * key field DEVNAME(32768) error: has a missing key
  * record's id-number=105 is out of range or inappropriate for its file
  * record's dba is inconsistent: [110:3014759]

Data file devices.dat: problems at page zero
  * invalid value 1273 for total records of record type DEVICES: actual=1901

Phase 1 of of consistency check: 3819 errors encountered in 1909 records of file devices.dat
Data file stores.dat: total 2 slots for records
Data file cart.dat: total 842 slots for records
Data file comp.dat: total 0 slots for records
Data file pools.dat: total 4 slots for records
Data file media.dat: total 714 slots for records
Data file mtopt.dat: total 2 slots for records
Data file mtopt.ext: total 0 slots for records
Data file devices.ext: total 38 slots for records
Data file stores.ext: total 6 slots for records

3819 errors encountered in 1909 records/nodes/pages

 

writedb fails on devices.dat.  Since I can't write the db to text then I can't do the writedb/readdb routine to fix the corruption.

Suggestions?

 

↧

(DP) Support tip: Experiencing performance problems with ECC turned on ?

February 8, 2016, 5:29 am
$
0
0

If you are experiencing performance problem with ECC turned on then please check the /var./opt/omni/log/debug.log on the cell server for any errors dealing with networking.   Also, you can check the /etc/opt/omni/server/cell/cell_info and /etc/opt/omni/server/config files on the cell server if UNIX machines.  Make sure that the same number of systems are in both files.  For Windows, check <OMNIBACK_DATA_HOME>\log\debug.log and for the ECC check, please <OMNIBACK_DATA_HOME>\Config\Server\cell\cell_info and <OMNIBACK_DATA_HOME>\Config\Server\Config to make sure these have the same number of systems in each.

↧
↧

(DP) Support Tip: Running rman manually to load the DataProtector Oracle library

February 8, 2016, 5:43 am
$
0
0

To manually test the loading of the DataProtector Oracle library with rman, please use the following example.  The example will also create DataProtector debug logs in /tmp which can viewed in case there are any problems loading the library :

RMAN> connect target sys/<password>@<SID>
RMAN> RUN {
2> ALLOCATE CHANNEL 'DP_TEST' TYPE 'SBT_TAPE'
3> PARMS 'SBT_LIBRARY=/opt/omni/lib/libob2oracle8_64bit.so,ENV=(OB2OPTS=-debug 1-500 ORA)';
4> RELEASE CHANNEL 'DP_TEST';

If using windows, the example would be the following and the debug would go into <OMNIBACK_DATA_HOME>\tmp :

RMAN> connect target sys/<password>@<SID>
RMAN> RUN {
2> ALLOCATE CHANNEL 'DP_TEST' TYPE 'SBT_TAPE'
3> PARMS 'SBT_LIBRARY=C:\PROGRA~1\OmniBack\bin\orasbt.dll,ENV=(OB2OPTS=-debug 1-500 ORA)';
4> RELEASE CHANNEL 'DP_TEST';

↧

(DP) Support Tip: EADR on Linux - Collecting debugs

February 8, 2016, 6:21 am
$
0
0

With EADR on Linux, in order to be able to collect / extract debugs from the machine, an nfsmount point needs to be available, so teh debugs can be redirected to that area for collection / extraction.

Once prompted with the main options, select option 6 to exit to shell.  Confirm that an IP ADDRESS is bound to the network interface. (ipconfig -a) ...then mount the share.  Example: mount 1.1.1.1:/nfsshare /mnt -o mountvers=3

Once this is done, type exit, to go back to main menu and select recovery option....when prompted, press the space bar to abort recovery, then select option "4" - Start recovery process, then select option "5" Use Debugs, finally, Option "2" Specify different debug option.

Example: -debug 1-200 /mnt/HPDBG.TXT 

 

↧
Search

All of a sudden drive choice errors

February 8, 2016, 4:14 pm
$
0
0

We have our backups set to use any of 14 virtual drives.  Always have.  So when a copy uses some of the drives and tape drives (for example VLS drives 10 and 11 and LTO drives 3 and 4), a backup that is set to use any drive 1-14 can pick ANY number that is open.  TODAY the software is giving the false alarm:

[Critical] From: BMA@sapbck.ssmhc.com "HP:Ultrium 3-SCSI_10_sapbck" Time: 2/8/2016 5:56:09 PM
[90:63] => UMA@sapbck.ssmhc.com@/dev/rchgr/autoch1
Cannot load exchanger medium (Target drive is busy.)

Duh! it is being used and the backup can choose from 12 other drives!  And it does BUT I get the error now.  How do I stop DP 8.1 from a false alarm on EVERY backup!  About the ONLY change that was made was the server the CM is on got bounced.

Running DP 8.1 on HP-UX 11.31

↧

which is the fastest way of backing up to disk

February 9, 2016, 1:03 am
$
0
0

I've been using a file library to do backups to a server, but I’ve also run the same job to a tape library. The backup to the tape library completed in under 2 hours, but the file library backup took 4 hours.

These jobs were using the same media host, and i accidently run both last night on the same set of files at the same time. The file library is a EVA LUN presented to the media, and the tape library is a Quatum iScaler 2000 which has been zoned to the same media host.

The main reason i could see that the tape backup completed quicker is due to the fact that you're not supposed to use concurrency on file library’s. So in this job I’m backup up 6 servers, which has 10 directories over these 6 servers. its only ever backing up 1 directory at a time, where as to the tape library its steaming all 10 directories at once.

If I’m right by this, would changing the file library to a VTL run quicker? have i not set up file library right but only 1 concurrency?

advice greatly apreciated.

 

 

↧
↧

required simple Procedure document to configure sql db backups using hpdp

February 9, 2016, 5:41 am
$
0
0

Hi Folks, I need the simple procedure documents of sql db backups using hpdp and oracle db backups using hpdp.So please provide the procedure or document if you have.

Thanks in advace,

Regards,

Steve

↧

(DP) Support Tip: Backup Encrypted Control Communication ( ECC ) configuration files

February 9, 2016, 7:38 am
$
0
0

If the Encrypted Control Communication ( ECC ) configuration files on the Cell Manager become corrupted for any reason, access to the clients will be lost.  ECC can manually be disabled but if there are a large number of clients, this can be a time consuming chore. It may make sense to keep a copy of the necessary files in the event corruption should occur. Hopefully you will never need them but if you chose to backup the files, they can be found at the following locations.

　

cell_info file:

Windows: Data_Protector_program_data\Config\Server\cell\cell_info

HP-UX and Linux systems: /etc/opt/omni/server/cell/cell_info

　

config file in:

Window: Data_Protector_program_data\Config\Server\config

HP-UX and Linux systems: /etc/opt/omni/server/config

　

certificates in:

Windows: Data_Protector_program_data\Config\Server\certificates

HP-UX and Linux systems: /etc/opt/omni/server/certificates

↧

(DP) Support Tip:

February 9, 2016, 11:06 am
$
0
0

Data Protector product installs and patch installs generate debug trace logs by default.  On Windows machines the logs are written to the following path:

     C:\Users\<User executing installation>\AppData\Local\Temp

      The user temp directory is definable in the registry with the following keys:

      Key Name:        HKEY_USERS\.DEFAULT\Environment
      Class Name:      <NO CLASS>
 
      Value 0
      Name:           TEMP
      Type:            REG_EXPAND_SZ
      Data&colon;            %USERPROFILE%\AppData\Local\Temp

      Value 1
      Name:           TMP
      Type:            REG_EXPAND_SZ
      Data&colon;            %USERPROFILE%\AppData\Local\Temp

↧
Viewing all 3189 articles
Browse latest View live
Search
<script src="https://jsc.adskeeper.com/r/s/rssing.com.1596347.js" async> </script>