Re: DSM_LOG setting on RedHat 64 bit
Turns out the isuse was the errorlogname setting appearing in the dsm.sys, which over-rides the DM_LOG. Pete On 12/10/2007, Peter Hitchman <[EMAIL PROTECTED]> wrote: > > Hi, > I am trying to use the dsmc client as a normal user, but it will not use > the DSM_LOG env var to create the errorlog into an area into which I can > write, it keeps trying to use the default one owned by root. > > I found this note from IBM: > > "Setting the DSM_LOG environment variable on Linux86 does not work with > /etc/environment on SuSE and RedHat. Please use the correlating > configuration file on Linux86 to set the environment for HSM daemons." > > But my RedHat knowledge is not yet up to understanding what they are on > about, what is the "correlating configuration file"? > > Thanks > > Pete >
DSM_LOG setting on RedHat 64 bit
Hi, I am trying to use the dsmc client as a normal user, but it will not use the DSM_LOG env var to create the errorlog into an area into which I can write, it keeps trying to use the default one owned by root. I found this note from IBM: "Setting the DSM_LOG environment variable on Linux86 does not work with /etc/environment on SuSE and RedHat. Please use the correlating configuration file on Linux86 to set the environment for HSM daemons." But my RedHat knowledge is not yet up to understanding what they are on about, what is the "correlating configuration file"? Thanks Pete
Re: Selective backup of directories that are mounted file systems
Hi, Thanks from the reply. I just think in terms of backing everything up from a certain point downwards making sense, because the sub-directories should be related. But I take your point, from a file system point of view its cleaner not to just cross boundaries when the backup takes place. Thanks Pete On 01/03/07, Richard Sims <[EMAIL PROTECTED]> wrote: On Mar 1, 2007, at 5:15 AM, Peter Hitchman wrote: > Hi, > To answer my own question , I could specify more than pathname in the > "selective" command. > But what I am doing here is a cold database backup and I have more > than one > db to deal with and I did not want to have customized scripts to > deal with > each database, just because some of the sub-directories happen to > be used as > mount points for file systems. > > Is there a good explanation why TSM acts in this way? It's simply rational behavior. You never want backups straying into file systems other than you specified should participate in the operation. Consider the extreme example of performing a backup of '/', and you happen to have /afs mounted... You need to specify just what you want backed up by Selective, which can be performed most readily by a shell or perl script in the operating system environment. Richard Sims
Re: Selective backup of directories that are mounted file systems
Hi, To answer my own question , I could specify more than pathname in the "selective" command. But what I am doing here is a cold database backup and I have more than one db to deal with and I did not want to have customized scripts to deal with each database, just because some of the sub-directories happen to be used as mount points for file systems. Is there a good explanation why TSM acts in this way? Regards Pete
Selective backup of directories that are mounted file systems
Hi, TSM 5.3 I am doing a selective backup like this: dsmc selective /data/XX/ -optfile=/opt/tivoli/tsm/client/ba/bin/dsmnotdp.opt -subdir=yes but some of the "directories" under /data/XX are mounted file systems and they do not get backed up. This is the documented behaviour of the -subdir option (in LinuxBackup-Archive Clients Installation and User's Guide for Linux and Unix 5.3), but it does not say if it is possible to automatically include the directories that are file systems. Is it possible to get this done with one "selective" backup command? Thanks Pete
Monitoring a clients backup progress
Hi, I am running the dsmc 5.2 client on Solaris. From a client's point of view is there a way to monitor the progress of a backup to get some idea of how complete it is? Something like "x% done" would be nice. Regards Pete
Using preschedulecmd
Hi, Using a 5.3 client on Solaris. Is it possible to schedule a command which is a script that invokes the dsmc command and uses preschedulecmd. Can I use this as an option to the dsmc programme or do I put it in the dsm.sys? Thanks Pete
Re: TDP oracle to new domain policy - error deleting old backups from old domain
Hi, I believe what you said about the management classes is true. What rman errors are you getting? Did you change the nodename? Doing that upsets the combination of rman and tdp in finding backups. Regards Pete -Original Message- From: ADSM: Dist Stor Manager [mailto:[EMAIL PROTECTED] Behalf Of Tae Kim Sent: 30 January 2006 17:27 To: ADSM-L@VM.MARIST.EDU Subject: [ADSM-L] TDP oracle to new domain policy - error deleting old backups from old domain Got a question for you all. Just move a tdp for oracle node to a different domain policy. this domain policy does not have management class of the same name that the node was using for the backup from its old domain policy. It is my understanding that once the node gets moved to a new domain, it would look for management class of the same name that it has bind its backups to and use that management class. If that management class does not exists, it would use the default management class. Is this true? Also we are having issues with deleting older DB from RMAN. what can cause these issues? Also the default management class of the new policy domain has different retention time. Would it help if I created the same management class with same backup copy group as the old one in the new policy domain? TIA, Tae Here is the env. TSM 5.2.4.4 on AIX 5.2 ML5 TDP for Oracle 5.2 on AIX 5.2 ML5 config of the node Platform: TDP Oracle AIX Client OS Level: 5.2 Client Version: Version 5, Release 2, Level 2.5 Policy Domain Name: FI_FIRST_DOM Last Access Date/Time: 01/30/06 12:15:13 Days Since Last Access: <1 Password Set Date/Time: 03/23/04 13:45:27 Days Since Password Set: 678 Invalid Sign-on Count: 0 Locked?: No Contact: Compression: Client Archive Delete Allowed?: Yes Backup Delete Allowed?: Yes Registration Date/Time: 03/22/04 11:56:24 Registering Administrator: Last Communication Method Used: Tcp/Ip Bytes Received Last Session: 4.50 M Bytes Sent Last Session: 17,599 Duration of Last Session: 0.67 Pct. Idle Wait Last Session: 18.32 Pct. Comm. Wait Last Session: 60.21 Pct. Media Wait Last Session: 0.00 Optionset: URL: Node Type: Client Password Expiration Period: 9,999 Day(s) Keep Mount Point?: No Maximum Mount Points Allowed: 2 Auto Filespace Rename : No Validate Protocol: No TCP/IP Name: TCP/IP Address: Globally Unique ID: Transaction Group Max: 0 Data Write Path: ANY Data Read Path: ANY Session Initiation: ClientOrServer High-level Address: Low-level Address: +=+ This message may contain confidential and/or privileged information. If you are not the addressee or authorized to receive this for the addressee, you must not use, copy, disclose or take any action based on this message or any information herein. If you have received this message in error, please advise the sender immediately by reply e-mail and delete this message. Thank you for your cooperation. +=+
Re: Gene - syntax/punctuation for an object in a schedule
Hi, When you set these things up, remember that the command is being passed to a shell for execution, since most Unix shells do not like a space between a redirection operator and a filename, the command is probably failing. You should find the failure logged on the client in tsm home area. Regards Pete -Original Message- From: ADSM: Dist Stor Manager [mailto:[EMAIL PROTECTED] On Behalf Of John Monahan Sent: 19 January 2006 22:21 To: ADSM-L@VM.MARIST.EDU Subject: Re: [ADSM-L] Gene - syntax/punctuation for an object in a schedule The entire command should be surrounded in single quotes. I believe it will also help if you remove the space before backset.out like so: >backset.out What interface are you using to create the sched - Admin Center, older web admin, or command line? __ John Monahan Senior Consultant Enterprise Solutions Computech Resources, Inc. Office: 952-833-0930 ext 109 Cell: 952-221-6938 http://www.computechresources.com Gene Shaffer <[EMAIL PROTECTED]> Sent by: "ADSM: Dist Stor Manager" 01/19/2006 02:50 PM Please respond to "ADSM: Dist Stor Manager" To ADSM-L@VM.MARIST.EDU cc Subject Gene - syntax/punctuation for an object in a schedule Hi; I am creating a schedule that invokes a script which generates backupsets. The Action field is set to COMMAND and in the Object field I am trying to pass the string to execute the script. It is: nohup ./backupset/gen_backupset2.sh > backset.out I believe it needs single or double quotes or something around either the whole line or part of the line. I have tried numerous permutations, but when I update it with TODAY and NOW to get it to run and test it out it just sits there. Any ideas? Gene Shaffer Software Systems Engineer Aurora Health Care " Experience is something you don't get until just after you need it"
Running dsmcad to show information from secondary tsm nodename
Hi, To enable backups of Oracle 7 datafiles and to separate these out from the normal backups I registered the machine again using a different nodename. Is it possible to run the dsmcad programme in such a way that it will accept connections to the machine and show the files that have been backed up on that machine using the alternative nodename? Regards Pete
Re: Over-ride include exclude options in a dsmc session
Hi, Thanks for the feedback. Pete -Original Message- From: ADSM: Dist Stor Manager [mailto:[EMAIL PROTECTED] On Behalf Of Andrew Raibeck Sent: 29 November 2005 22:49 To: ADSM-L@VM.MARIST.EDU Subject: Re: [ADSM-L] Over-ride include exclude options in a dsmc session Here are some options you can think about, or you can create your own variations. I'm sure there are other methods, too. * Consider using PRESCHEDULECMD and POSTSCHEDULECMD to stop and start Oracle 7 while the regularly scheduled incremental backup occurs. * If the regular incremental backup schedule conflicts with when you can back up Oracle, see if yuo can find a common timeframe that works for both. * Bind the Oracle files to a management class whose copy group FREQUENCY is set to some nonzero value (7, for example). This will prevent the regular incremental backup from trying to back the files up if they were backed up earlier in a separate backup process for Oracle. Then create a second schedule to back up Oracle, with PRE/POSTSCHEDULECMD options as mentioned above, and make the schedule a SELECTIVE operation against the directory/files that contain the Oracle files. * Create a second node that is just for backing up the Oracle files, with a schedule that targets just the Oracle directory/files. Regards, Andy Andy Raibeck IBM Software Group Tivoli Storage Manager Client Development Internal Notes e-mail: Andrew Raibeck/Tucson/[EMAIL PROTECTED] Internet e-mail: [EMAIL PROTECTED] IBM Tivoli Storage Manager support web page: http://www-306.ibm.com/software/sysmgmt/products/support/IBMTivoliStorag eManager.html The only dumb question is the one that goes unasked. The command line is your friend. "Good enough" is the enemy of excellence. "ADSM: Dist Stor Manager" wrote on 2005-11-29 15:14:47: > Thanks Andy, > My situation is that I have Oracle 7 databases to backup, so no TDP > and rman. The system is set so that the areas we store the Oracle > datafiles in is ignored. Fine for the Unix backups, but I need to be > able to at least shutdown the Oracle 7 databases and do a cold backup. > At the same time I do not want to change the inclexcl file named in > the dsm.sys, its no good backing up the database while its running > (well I know there is the hot tablespace backup option, but lets > ignore that for now). Also I need to back these files up to the same > backup policy used for the Unix backup, not the one used for rman > backups, because that policy is set to never automatically age out the > files. > > Regards > > Pete > > -Original Message- > From: ADSM: Dist Stor Manager [mailto:[EMAIL PROTECTED] On Behalf > Of Andrew Raibeck > Sent: 29 November 2005 21:32 > To: ADSM-L@VM.MARIST.EDU > Subject: Re: [ADSM-L] Over-ride include exclude options in a dsmc > session > > > No, but your TSM administrator should be able to configure an > include-exclude file in dsm.sys (see the INCLEXCL option in the client > manual). > > Considering the ramifications of modifying Include-exclude options, > these should be treated as semi-permanent and not changed unless you > are looking to change the policy by which your files are managed. > > If this does not answer your question, then please clarify what you > need to do. > > Regards, > > Andy > > Andy Raibeck > IBM Software Group > Tivoli Storage Manager Client Development > Internal Notes e-mail: Andrew Raibeck/Tucson/[EMAIL PROTECTED] Internet > e-mail: [EMAIL PROTECTED] > > IBM Tivoli Storage Manager support web page: > http://www-306.ibm.com/software/sysmgmt/products/support/IBMTivoliStor > ag > eManager.html > > The only dumb question is the one that goes unasked. > The command line is your friend. > "Good enough" is the enemy of excellence. > > "ADSM: Dist Stor Manager" wrote on 2005-11-29 > 14:11:30:
Re: Over-ride include exclude options in a dsmc session
Thanks Andy, My situation is that I have Oracle 7 databases to backup, so no TDP and rman. The system is set so that the areas we store the Oracle datafiles in is ignored. Fine for the Unix backups, but I need to be able to at least shutdown the Oracle 7 databases and do a cold backup. At the same time I do not want to change the inclexcl file named in the dsm.sys, its no good backing up the database while its running (well I know there is the hot tablespace backup option, but lets ignore that for now). Also I need to back these files up to the same backup policy used for the Unix backup, not the one used for rman backups, because that policy is set to never automatically age out the files. Regards Pete -Original Message- From: ADSM: Dist Stor Manager [mailto:[EMAIL PROTECTED] On Behalf Of Andrew Raibeck Sent: 29 November 2005 21:32 To: ADSM-L@VM.MARIST.EDU Subject: Re: [ADSM-L] Over-ride include exclude options in a dsmc session No, but your TSM administrator should be able to configure an include-exclude file in dsm.sys (see the INCLEXCL option in the client manual). Considering the ramifications of modifying Include-exclude options, these should be treated as semi-permanent and not changed unless you are looking to change the policy by which your files are managed. If this does not answer your question, then please clarify what you need to do. Regards, Andy Andy Raibeck IBM Software Group Tivoli Storage Manager Client Development Internal Notes e-mail: Andrew Raibeck/Tucson/[EMAIL PROTECTED] Internet e-mail: [EMAIL PROTECTED] IBM Tivoli Storage Manager support web page: http://www-306.ibm.com/software/sysmgmt/products/support/IBMTivoliStorag eManager.html The only dumb question is the one that goes unasked. The command line is your friend. "Good enough" is the enemy of excellence. "ADSM: Dist Stor Manager" wrote on 2005-11-29 14:11:30:
Over-ride include exclude options in a dsmc session
Hi, I am running a tsm client on Solaris. When I execute the dsmc command to backup a group of directories and the files they contain is there a way to include pathnames that have been excluded by an exclude entry in a file specified by the inclexcl option? Thanks Pete
Re: Antwort: query session command - for a specific session
Thanks for the replies. It now becomes clear, I thought I was looking at two numbers separated by a comma, but its just the local locale way of formatting a number into thousands. Regards Pete -Original Message- From: ADSM: Dist Stor Manager [mailto:[EMAIL PROTECTED] Behalf Of Richard Sims Sent: 28 October 2005 13:55 To: ADSM-L@VM.MARIST.EDU Subject: Re: [ADSM-L] Antwort: query session command - for a specific session On Oct 28, 2005, at 7:43 AM, Alexander Verkooijen wrote: > Just nitpicking but "show session {xxx}" > seems to ignore the session number. > It displays information for all session > regardless whether a sessions number is > provided or not. > Indeed, it does not take an operand. The developers provide us the opportunity to get the info we need via a Select command on the SESSIONS table. The original poster should note that number and date formats are governed by locale; and numbers entered in commands should be simple integer sequences, not in the formatted output form as displayed in response to commands. Richard Sims
query session command - for a specific session
Hi, Running Version 5, Release 2, Level 3.4. Something that has been bugging me for a while. The "query session" command shows all of the tsm sessions and according to the help you can narrow this down to a specific session using "query session 'sesnum'". But I can never get this to work. The output I see looks like the example below and the session number column format is not what it looks like in the docs:- Sess Comm. Sess Wait Bytes Bytes Sess Platform Client Name Number Method StateTimeSent Recvd Type --- -- -- -- --- --- - --- 129,541 Tcp/Ip MediaW 4.1 M1.4 K 1.0 K Node TDP Ora- x How can I run a detailed query command for just this session? Regards Pete
Re: Scheduled session remains active
Hi, My exprience is that it is the dsmc programme that runs the scheduled jobs, usually when things get "odd" we restart this scheduler process on the client. dsmcad is the deamon required to run the web interface to do backup and restores. But I am on Solaris not RD Linux. Regards Pete -Original Message- From: ADSM: Dist Stor Manager [mailto:[EMAIL PROTECTED] Behalf Of Jurjen Oskam Sent: 15 September 2005 11:48 To: ADSM-L@VM.MARIST.EDU Subject: [ADSM-L] Scheduled session remains active Hi, I have a problem with a scheduled command. The TSM server is 5.3.1.4 on AIX 5.3 ML2, the affected client is 5.2.2.0 on Red Hat Linux Advanced Server 2.1 (i386). SCHEDmode is PROMPTed, the client has MANAGEDSERVICES WEBCLIENT SCHEDULE and dsmcad is runnning. There are two schedules defined for this client, to run after each other: Policy Domain Name: PD_TRIAS Schedule Name: DB_DIGIDATA_DDM+MIS Description: Backup DigiData databases Action: Command Options: Objects: /home/oracle/scripts/sh/backup-digiddm-digimis.sh Priority: 5 Start Date/Time: 02/25/05 21:00:00 Duration: 30 Minute(s) Schedule Style: Classic Period: 1 Day(s) Day of Week: Weekday Month: Day of Month: Week of Month: Expiration: Last Update by (administrator): JOSKAM Last Update Date/Time: 09/14/05 11:19:42 Managing profile: Policy Domain Name: PD_TRIAS Schedule Name: INCR_AIX Description: Daily incremental backup for AIX systems Action: Incremental Options: Objects: Priority: 5 Start Date/Time: 04/12/02 22:00:00 Duration: 4 Hour(s) Schedule Style: Classic Period: 1 Day(s) Day of Week: Weekday Month: Day of Month: Week of Month: Expiration: Last Update by (administrator): JOSKAM Last Update Date/Time: 09/14/05 11:34:31 Managing profile: As you can see, the first schedule completed successfully and on time: Policy Domain Name: PD_TRIAS Schedule Name: DB_DIGIDATA_DDM+MIS Node Name: COOKIEMONSTER Scheduled Start: 09/14/05 21:00:00 Actual Start: 09/14/05 21:00:09 Completed: 09/14/05 21:42:13 Status: Completed Result: 0 Reason: All operations completed successfully. The problem is that the second schedule (INCR_AIX) is never started. SCHEDMODE is PROMPTed, but the TSM server never prompts the client to start the INCR_AIX schedule. I believe I have found the cause for this: 09/14/05 21:00:01 ANR2561I Schedule prompter contacting COOKIEMONSTER (session 2636) to start a scheduled operation. (SESSION: 26) 09/14/05 21:00:09 ANR0406I Session 2637 started for node COOKIEMONSTER (Linux86) (Tcp/Ip cookiemonster.trias.nl(33983)). (SESSION: 2637) [...] 09/14/05 21:42:13 ANR2507I Schedule DB_DIGIDATA_DDM+MIS for domain PD_TRIAS started at 09/14/2005 21:00:00 for node COOKIEMONSTER completed successfully at 09/14/2005 21:42:13. (SESSION: 2637) 09/14/05 21:42:13 ANR0403I Session 2637 ended for node COOKIEMONSTER (Linux86). (SESSION: 2637) 09/14/05 21:42:13 ANR0406I Session 2700 started for node COOKIEMONSTER (Linux86) (Tcp/Ip cookiemonster.trias.nl(34037)). (SESSION: 2700) 09/14/05 21:42:13 ANR0403I Session 2700 ended for node COOKIEMONSTER (Linux86). (SESSION: 2700) 09/15/05 02:00:01 ANR2578W Schedule INCR_AIX in domain PD_TRIAS for node COOKIEMONSTER has missed its scheduled start up window. 09/15/05 03:00:34 ANR0403I Session 2636 ended for node COOKIEMONSTER (). (SESSION: 26) As you can see, session 2636 isn't ended until 03:00:34. I don't understand why. My guess is that it's ended due to a timeout on the server side (COMMTIMEOUT is 3600). On the client, the scheduler logfile contains: [...] 09/14/2005 21:00:09 Executing Operating System command or script: /home/oracle/scripts/sh/backup-digiddm-digimis.sh 09/14/2005 21:42:13 Finished command. Return code is: 0 09/14/2005 21:42:13 ANS1908I The scheduled command completed successfully. 09/14/2005 21:42:13 Scheduled event 'DB_DIGIDATA_D
Re: Oracle TDP - Help with removing files from a filespace created by rman nocatalog
Hi, Thanks for the reply. Starting again is the conclusion I have arrived at, as well. Point taken about how you backup the catalog db, I should consider how I could schedule this in. Thanks Pete -Original Message- From: ADSM: Dist Stor Manager [mailto:[EMAIL PROTECTED] Behalf Of Andreas Almroth Sent: 08 September 2005 14:06 To: ADSM-L@VM.MARIST.EDU Subject: Re: [ADSM-L] Oracle TDP - Help with removing files from a filespace created by rman nocatalog Peter Hitchman wrote: >Hi, >I have backups using TDP for Oracle of my rman catalog database. These are >done in nocatalog mode, so the records of the backup are kept in the database >control file. My problem is that I now have files in a filespace that are no >longer referenced by rman and I do not know how to delete them. What is my >best way of getting rid of these backup entries in TSM? > >Regards > >Pete > > Tricky one. OK, I must admit I'm not a wizard, but nothing immediate comes to mind, apart from deleting the filespace for the client node, and start all over. That will of course delete any valid backup pieces as well, which is bad, but you could make a new backup manually after deleting the filespace, and you would have at least one backup again. I have choosen to backup the catalog db in cold mode, just using the normal dsm scheduler, shutting down and starting the instance as part of the pre and post backup scripts, and thus not involve the TDP module at all. This way I don't need a working RMAN/TDP environment if I ever need restore the catalog instance. /Andreas
Re: Temporary disabling a client scheduled job
Hi, Thanks I'll give that a go. Pete -Original Message- From: ADSM: Dist Stor Manager [mailto:[EMAIL PROTECTED] Behalf Of Bill Dourado Sent: 05 September 2005 10:48 To: ADSM-L@VM.MARIST.EDU Subject: Re: [ADSM-L] Temporary disabling a client scheduled job 3) Use EXPIRATION field , enter a date. Bill Zoltan Forray/AC/VCU <[EMAIL PROTECTED]> Sent by: "ADSM: Dist Stor Manager" 02/09/2005 16:24 Please respond to "ADSM: Dist Stor Manager" To: ADSM-L@VM.MARIST.EDU cc: Subject:Re: [ADSM-L] Temporary disabling a client scheduled job Two ways that I can think of: 1. Dis-Associate the client/node from the schedule (effects one or more nodes in this schedule) 2. Change the STARTDATE to a date in the future (effects ALL nodes in this schedule) Peter Hitchman <[EMAIL PROTECTED]> Sent by: "ADSM: Dist Stor Manager" 09/02/2005 11:04 AM Please respond to "ADSM: Dist Stor Manager" To ADSM-L@VM.MARIST.EDU cc Subject [ADSM-L] Temporary disabling a client scheduled job Hi, Using TSM 5.2. I have a set of client schedules defined, is it possible for any one of them to temporarily be disabled? I could not see anything in the description of the update schedule command to do this. I could of course do a delete to stop a job running, but it would be better if I could just disable it. Regards Pete *** Disclaimer *** This email and any files transmitted with it contains privileged and confidential information and is intended for the addressee only. If you are not the intended recipient you must not disseminate, distribute or copy this email. Please notify the sender at once that you have received it by mistake and then delete it from your system. Any views expressed in this email are those of the sender and not necessarily those of ALSTEC or its associated companies. Whereas ALSTEC takes reasonable precautions to ensure this email is virus free, no responsibility will be accepted by it for any loss or damage whether direct or indirect which results from the contents of this email or its attachments. ALSTEC Group Limited - Registered in England No.3939840 - www.alstec.com Registered office: Cambridge Road, Whetstone, Leicester. LE8 6LH. England. Telephone: +44 (0)116 275 0750
Oracle TDP - Help with removing files from a filespace created by rman nocatalog
Hi, I have backups using TDP for Oracle of my rman catalog database. These are done in nocatalog mode, so the records of the backup are kept in the database control file. My problem is that I now have files in a filespace that are no longer referenced by rman and I do not know how to delete them. What is my best way of getting rid of these backup entries in TSM? Regards Pete
Temporary disabling a client scheduled job
Hi, Using TSM 5.2. I have a set of client schedules defined, is it possible for any one of them to temporarily be disabled? I could not see anything in the description of the update schedule command to do this. I could of course do a delete to stop a job running, but it would be better if I could just disable it. Regards Pete
Re: TDP for Oracle 8.1.7.4 & 9.2.0.6 on same server
Hi, I know you said the links are OK, but this type of error usually indicates a problem with the $ORACLE_HOME/lib/libobk.so symbolic link. I cannot find anything for 585 SBT error code. Other possibilities could be an formatting error in the parms part of the channel allocation or permissions to read config files such as dms.sys. Good luck Pete -Original Message- From: ADSM: Dist Stor Manager [mailto:[EMAIL PROTECTED] On Behalf Of Scott Grover Sent: 01 September 2005 21:51 To: ADSM-L@VM.MARIST.EDU Subject: [ADSM-L] TDP for Oracle 8.1.7.4 & 9.2.0.6 on same server Hi all, I have TDP 64 bit for Oracle AIX 5.2.0 running on an AIX 5.2 client. I believe I have the symbolic links for both Oracle instances set up properly, and have done "relink all" for both successfully. I also have permissions set to 666 on the tdpoerror logs, etc. - but getting the following error on the Oracle 8.1.7 instance (Oracle 9 runs fine): RMAN-03007: retryable error occurred during execution of command: allocate RMAN-07004: unhandled exception during command execution on channel ch00 RMAN-10035: exception raised in RPC: ORA-19554: error allocating device, device type: SBT_TAPE, device name: ORA-19557: device error, device type: SBT_TAPE, device name: ORA-27000: skgfqsbi: failed to initialize storage subsystem (SBT) layer IBM AIX RISC System/6000 Error: 14: Bad address Additional information: 585 ORA-19511: Unknown SBT error code = 585, errno = 14 RMAN-10031: ORA-19624 occurred during call to DBMS_BACKUP_RESTORE.DEVICEALLOCATE Has anyone else run into this? Thanks in advance if anyone has any ideas ! Here's showenv output for both oracle instances; ts90 is 64 bit and ts91 is 32 bit: TS90: svho1ds612 # tdpoconf showenv -tdpo_optfile=/usr/tivoli/tsm/client/oracle/bin64/tdpo_ts90.opt IBM Tivoli Storage Manager for Databases: Data Protection for Oracle Version 5, Release 2, Level 0.0 (C) Copyright IBM Corporation 1997, 2003. All rights reserved. DATA PROTECTION FOR ORACLE INFORMATION Version: 5 Release: 2 Level:0 Sublevel: 0 Platform: 64bit TDP Oracle AIX TSM SERVER INFORMATION Server Name: TDPO_TS90_DS612 Server Address: SVHO1TSM1.SUPERVALU.COM Server Type: AIX-RS/6000 Server Port: 1500 Communication Method: TCP/IP SESSION INFORMATION Owner Name: Node Name:tdpo_ts90_ds612 Node Type:TDP Oracle AIX DSMI_DIR: /usr/tivoli/tsm/client/api/bin64 DSMI_ORC_CONFIG: /usr/tivoli/tsm/client/oracle/bin64/dsmd_ts90.opt TDPO_OPTFILE: /usr/tivoli/tsm/client/oracle/bin64/tdpo_ts90.opt Password Directory: /etc/adsm Compression: FALSE TS91: svho1ds612 # tdpoconf showenv -tdpo_optfile=/usr/tivoli/tsm/client/oracle/bin64/tdpo_ts91.opt IBM Tivoli Storage Manager for Databases: Data Protection for Oracle Version 5, Release 2, Level 0.0 (C) Copyright IBM Corporation 1997, 2003. All rights reserved. DATA PROTECTION FOR ORACLE INFORMATION Version: 5 Release: 2 Level:0 Sublevel: 0 Platform: 64bit TDP Oracle AIX TSM SERVER INFORMATION Server Name: TDPO_TS91_DS612 Server Address: SVHO1TSM1.SUPERVALU.COM Server Type: AIX-RS/6000 Server Port: 1500 Communication Method: TCP/IP SESSION INFORMATION Owner Name: oracle Node Name:tdpo_ts91_ds612 Node Type:TDP Oracle AIX DSMI_DIR: /usr/tivoli/tsm/client/api/bin64 DSMI_ORC_CONFIG: /usr/tivoli/tsm/client/oracle/bin64/dsmd_ts91.opt TDPO_OPTFILE: /usr/tivoli/tsm/client/oracle/bin64/tdpo_ts91.opt Password Directory: /etc/adsm Compression: FALSE svho1ds612 #
Re: TDPO - Oracle Rman Question
Hi, Your problem is that the backup stores with it the node it was backed up on and it only wants to restore to a machine that specifies that node name in the TDPO_OPTFILE being used. The easiest thing to IMHO is to NFS mount the file systems across from your test machine to the original machine and use the "set newname" rman command to change the path of the datafiles and then run the duplicate rman session from the original host machine. Of course this will not work for you if want the pathnames to be the same, but I assume that at least you are renaming the database and therefore will want the new database name if the pathnames. Also depending on the network situation it could be a bit slow 8->. Do you have MetaLink access? Oracle publish a note (265595.1) on what you have to do with the TDPO set-up to do the duplicate on the "other" host. Regards Pete [END] -Original Message- From: ADSM: Dist Stor Manager [mailto:[EMAIL PROTECTED] Behalf Of Muthukumar Kannaiyan Sent: 01 August 2005 18:59 To: ADSM-L@VM.MARIST.EDU Subject: [ADSM-L] TDPO - Oracle Rman Question Hello All, We did backup a oracle instance from our production machine using TDPO and RMAN. We are trying to restore this oracle instance on our test machine using oracle rman "duplicate" command. We are getting following error RMAN-03002: failure of Duplicate Db command at 07/27/2005 20:31:50 RMAN-03015: error occurred in stored script Memory Script ORA-19507: failed to retrieve sequential file, handle="f7gqisvp_1_1", parms="" ORA-27029: skgfrtrv: sbtrestore returned error ORA-19511: Error received from media manager layer, error text: ANS1302E (RC2)No objects on server match query RMAN> Do we have something similar to TSM client "set Access" command on TDPO client? TIA Muthu 202-458-8340 - Work
Backups using TDP for Oracle run twice
Hi, I am using TSM 5.2 and TDP for Oracle 5.2, on Solaris 8 The backups are run through a scheduler set-up for Oracle. On one particular box the backups are run twice, seconds apart. The scheduler log only records running the job once. Has anyone seen this kind of thing before? Thanks Pete