Tsm & TEC trouble
Hello, I configure TSM integration with TEC, all work fine, but episodically TSM stop sending events (without any error messages about this). This situation occurred two times: 1. TSM 5.2.2.0 and TEC 3.9FP02 on same mashine with RH 7.3. (After restart TSM all work) 2. TSM 5.2.2.0(or 5.2.3.0, I can't remember) on Windows 2000 SP4 and TEC 3.9FP02 on RH 7.3 Any ideas? Thanks
AW: How to view the Policy Information on the CLI ?
YEP :-P, thank you Andy Regards Joachim -Ursprüngliche Nachricht- Von: ADSM: Dist Stor Manager [mailto:[EMAIL PROTECTED] Auftrag von Andrew Raibeck Gesendet: Freitag, 28. Januar 2005 21:13 An: ADSM-L@VM.MARIST.EDU Betreff: Re: How to view the Policy Information on the CLI ? > You can get some of if with QUERY MGMTCLASS, but it only includes > the mgmtclasses available, not the copy group version info. For copy group detail, add the -DETAIL option :-) dsmc query mgmtclass -detail Regards, Andy
Backup Windows 2003 AD
Hi, since we moved our active directory environment from windows 2000 to windows 2003 we see a backup error with D:\windows\ntds\temp.edb. File is open and can not be saved The file is described somewhere in the web as a scratchpad for AD. I wonder if the file is needed for recovery. If not - why there is not standard tsm exclude? If yes - who should i backup the file? Kind regards Stefan Holzwarth
clear the drive path
Hi everbody, does any body know how to "clear the drive path" on any IBM library. Thanks
Re: Changing the profile
On Jan 30, 2005, at 9:48 PM, Hashim, Shukrie BSP-ISM/116 wrote: I need some help here, I have some data sitting on some tapes and I need to change the profile or the characteristics of the data to be ARCHIVED data can I do that ? is there any database manupulation .. involved ? thanks No. The essence of the TSM product is data integrity and preservation, as stored. You cannot meddle with the data from the standpoint of the server. All data alteration decisions are done at the client end, where the data is owned and where its purpose is defined. Presuming that the data you are writing about is Backup type, you could restore it to a special directory and from there perform an Archive, then Expire the backup data if you wish. Richard Simshttp://people.bu.edu/rbs
Re: clear the drive path
May be, remove path? -Original Message- From: ADSM: Dist Stor Manager [mailto:[EMAIL PROTECTED] On Behalf Of Moh Jar Sent: Monday, January 31, 2005 2:48 PM To: ADSM-L@VM.MARIST.EDU Subject: clear the drive path Hi everbody, does any body know how to "clear the drive path" on any IBM library. Thanks
Antwort: Re: clear the drive path
"delete path" is this what your looking for??? Christian
Re: Antwort: Re: clear the drive path
this request comes from the library console as a suggested solution to a drive failer. the retension period of the dismount have expired. the drive is locked now and I can not open the magazines of the libary normally. It says "the library in Loaked". On Mon, 31 Jan 2005 13:25:01 +0100, Christian Demnitz <[EMAIL PROTECTED]> wrote: > "delete path" > > is this what your looking for??? > > Christian >
Re: File has a corrupt named stream
It appears that you're trying to restore files from one OS (Macintosh) onto a machine running another OS (Windows). With a rare exception or two, that will not work, since the filesystem structures are radically different. -- Mark Stapleton ([EMAIL PROTECTED]) Berbee Information Networks Office 262.521.5627 >-Original Message- >From: ADSM: Dist Stor Manager [mailto:[EMAIL PROTECTED] On >Behalf Of Sujay Dinakar R. >Sent: Sunday, January 30, 2005 11:40 PM >To: ADSM-L@VM.MARIST.EDU >Subject: Re: File has a corrupt named stream > >Thanks Richards, > >Message ANS2048E, is comming the error log, > >Environment: > >Tsm server V. 5.1.0 and client V. 5.2.0 is on Windows 2000 Server. > >I was trying to restore the files to the Network path to a X >Server G5 Mac OS 10.3, when I try to restore the same files on >the local/Network W2K it goes smooth without any err. is there >any cross platform issue in restoring? I would like to know >this because I like to take the backup from the X raid which >is connected to Xserver from a Win 2K client, since all of my >previous backups are taken from the windows I want to continue >to user Win 2K client for backup/restore. > >Sujay Dinakar R. > > >Richard Sims <[EMAIL PROTECTED]> wrote: >On Jan 29, 2005, at 6:14 AM, Sujay Dinakar R. wrote: >> for all the files what ever I try to restore. >> To day I had received a problem in the TSM client: it is showing as >> "File has a corrupt named stream" >> >> Can any one please guide me why the problem is occuring. > >When posting, help us to help you by providing a realistic >amount of information, including your operating environment >type, TSM client/server levels, and file system type, as well >as message numbers (most typically found in the dsmerror.log >file). In this case, it would also be helpful to have a lot >more context, as in whether the restoral client and file >system are the same ones which originated the backup file you >are attempting to restore, and whether the operating system >environment had changed over the intervening interval. > >I would surmise that your environment is Windows and that the >file system type is NTFS. You may be trying to employ a >relatively modern TSM client to restore data which had been >backed up with an older client, where that backup client had a >defect such that in the presence of an unusual circumstance, >the data was not backed up correctly. This may be message >ANS2048E, where APAR IC33922 notes that the backup may be bad, >as addressed by other APARs. If this is the case, there isn't >much you can do except try to make the most of what data may >have restored. > >Richard Sims http://people.bu.edu/rbs > > >- >Do you Yahoo!? > The all-new My Yahoo! Get yours free! >
Re: Select statement for domains not giving correct output ??
Does your script do any filtering or other "massaging" of the data? If so, it could be that the one domain is being massaged out. Do you have more than one TSM server? If so, it is possible that the script is logging into a different server. It would help to see: - The script - The *complete* output from a manual SELECT - The *complete* output from the script (don't use -DATAONLY=YES and don't strip off any non-data messages... let's see the whole output) 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] 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 01/29/2005 17:04:00: > Hi, > > I have a script running daily on a linux host to query for the > status on backupjobs performed - but something wierd has happened > within the last month or so... When doing a query in the script like: > > SELECT * FROM DOMAINS > > I get a list og 9 domains But when doing the same query manually > (either via dsmc or on the command-line on the web-interface), the > list has 10 policy domains ?? It is the exact same user-id/password, > that is being used for the queries. Somehow policy-domain number 5 > (in the list) is missing when the query is performed from the script... Why ?? > > The server is running 5.2.2.5 (on a Windows 2000 Server) . > > Regards, > Brian Ipsen > > > --- > This mail was scanned for virus and spam by Progressive IT A/S. > ---
HSM Problems
Has anyone seen the following errors ? 01/29/05 11:29:32 ANS9178E dsmscoutd: cannot open file /dev/fsm: The requested resource is busy.. 01/29/05 11:29:32 ANS9178E dsmscoutd: cannot open file /dev/fsm: The requested resource is busy.. 01/29/05 11:29:32 ANS9178E dsmscoutd: cannot open file /dev/fsm: The requested resource is busy.. 01/29/05 11:29:32 ANS9178E dsmscoutd: cannot open file /dev/fsm: The requested resource is busy.. It seems as each time this error get logged another dsmscoutd process starts until thousands of scout procs are running and the system hangs. Thanks, Mark.
Re: required HBA fileset to use San Discovery for TSM 5.3
Thanks Dave, I Thought you needed to be at ML2 for Aix 5.2? to install TSM 5.3 although I guess it doesn't matter because we are at ML4 so we should be ok for TSM 5.3 and the Admin Center. Dave Canan wrote: > I agree, this is confusing. Let me see if I can list the requirements > here: > > AIX 5.2, 32-bit - devices.common.IBM.fc.hba-api 5.2.0.30 or higher > AIX 5.2 64-bit - devices.common.IBM.fc.hba-api > 5.2.0.30 (ML4 > or higher) > > AIX 5.3 32-bit - devices.common.IBM.fc.hba-api 5.3.0.0 (base level) > AIX 5.3 64-bit - devices.common.IBM.fc.hba-api 5.3.0.10 > > Also, in answer to your question regarding AIX 5.2 level for TSM > 5.3, you must be at release level 52f, which is the same as maintenance > level 3. > > At 01:06 PM 1/28/2005 -0500, you wrote: > >Does anyone know the correct fileset required to use San Discovery for > >TSM V5.3? > >(devices.common.IBM.fc.hba-api required) for AIX 5.2. > >I read in the TSM supported HBAs for San Device maping > >technote that it's 5.2.0.50. > > > >But I read in the ITSM,server & Storage Agent v5.3.0 readme > >where it said later than 5.2.0.30 for 64-bit. We are at 5.2.0.40. > > > >I also understand that 5.2.0.30 for 64-bit. Has the error. > >Does anyone know where else I can look on the IBM site > >to check for any errors on filesets? > > > >Also, Just to be sure what maintenance level does TSM 5.3 > >require for Aix 5.2. I believe it's ML2 we are at ML 4 so > >I assume we are ok and no apar is needed. > > > >Thanks for any help in advance > > > >Config > > > >TSM 5.2.3.1 & 5.2.3.5 > >AIX 5.2 Maintenance level 4 > >pseries 670 > > Dave Canan > TSM Performance > IBM Advanced Technical Support > [EMAIL PROTECTED]
TSM 5.1.6.3 server : Supported Windows Platforms?
Hi Guys, Can anyone point me towards a list of supported Windows platforms for TSM 5.1.6.3 ... Or tell me if its supported on Windows 2003 Server??? I have a migration to do... Can't find it on the web!?!?! Many Thanks Tony This e-mail and any files transmitted with it, are confidential and intended solely for the use of the addressee. The content of this e-mail may have been changed without the consent of the originator. The information supplied must be viewed in this context. If you have received this e-mail in error please notify our Helpdesk by telephone on +44 (0) 20-7444-8444. Any use, dissemination, forwarding, printing, or copying of this e-mail or its attachments is strictly prohibited.
TSM client for win98?
Hello All! I can't find subj. I have download package from partnerworld (In description writed about support win98), but package include only win2000 clients. Which version of client needed? Where I can find him? (Used TSM Server 5.2) Thanks
NEXTSTGPOOL with MAXSIZE does not work with MOVE DATA
We have a Storage Pool hierarchy something like: DISK Stgpool (noliimt) --> File Stgpool (2GB limit) --> Tape Stgpool (nolimit) Clients backup to DISK, files smaller than 2GB migrate to "File" stgpool, everything else migrates to Tape. We've changed the MAXSIZE to 2GB and wanted to move all files from Tape that are smaller than 2GB to the File Stgpool. This can be done by moving everything to DISK and let migration handle it. I thought it would be more efficient to just move this data to the File stgpool but ran into the following error: ANR1150W Move data process terminated for volume VOLNAME - unable to move file to target storage pool due to exclusion by size. So I thought it was weird that Migration will honor the Nextstgpool when a maxsize is set but MOVE DATA does not. (Also client backups/archives will honor the nextstgpool). Along with MOVE DATA, MOVE NODEDATA and RECLAMATION also do not work in this scenario. I called Tivoli Support and they said working as designed and are going to update the documentation to state this. My questions are: - does anyone else think this is working a bit weird (ie. Migration and client sessions work fine when writing to a stgpool with a maxsize and nextstgpool but move data, move nodedata and reclamation do not). - Any other (efficient) way of doing what I want - move the small files from tape to the file pool Thanks, Tim Rushforth City of Winnipeg
Re: TSM 5.1.6.3 server : Supported Windows Platforms?
The TSM support page at http://www-306.ibm.com/software/sysmgmt/products/support/IBMTivoliStorageManager.html contains links to server and client support information. See the "Server and Client Requirements" links on that page (hard to spot, but roughly midway down and in the center of the page). 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] 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 2005-01-31 08:44:27: > Hi Guys, > > Can anyone point me towards a list of supported Windows platforms for TSM > 5.1.6.3 ... > > Or tell me if its supported on Windows 2003 Server??? > > I have a migration to do... Can't find it on the web!?!?! > > Many Thanks > > Tony > > > > This e-mail and any files transmitted with it, are confidential and > intended solely for the use of the addressee. The content of this > e-mail may have been changed without the consent of the originator. > The information supplied must be viewed in this context. If you have > received this e-mail in error please notify our Helpdesk by > telephone on +44 (0) 20-7444-8444. Any use, dissemination, > forwarding, printing, or copying of this e-mail or its attachments is > strictly prohibited.
Re: TSM 5.1.6.3 server : Supported Windows Platforms?
>Can anyone point me towards a list of supported Windows platforms for TSM >5.1.6.3 ... Start at the TSM Support page http://www-306.ibm.com/software/sysmgmt/products/support/IBMTivoliStorageManager.html and review Server Requirements. Richard Sims
Re: TSM client for win98?
Windows 98 is no longer supported with TSM. The last client version that suppoted Win98 was 4.2. You can still find the download at ftp://ftp.software.ibm.com/storage/tivoli-storage-management/patches/client/v4r2. Go to the URL, and you should be able to navigate the rest of the way to the last patch level for the Win98 client (version 4.2.3.2). Again, note that this version is out of support. 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] 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 2005-01-31 08:50:20: > Hello All! > > I can't find subj. I have download package from partnerworld (In > description writed about support win98), but package include only > win2000 clients. Which version of client needed? Where I can find him? > (Used TSM Server 5.2) > > Thanks
Re: TSM 5.1.6.3 server : Supported Windows Platforms?
Many Thanks Richard, Andy and anyone else who assisted. The old eyes are not what they used to be!!! Best regards Tony -Original Message- From: ADSM: Dist Stor Manager [mailto:[EMAIL PROTECTED] On Behalf Of Richard Sims Sent: 31 January 2005 16:00 To: ADSM-L@VM.MARIST.EDU Subject: Re: TSM 5.1.6.3 server : Supported Windows Platforms? >Can anyone point me towards a list of supported Windows platforms for TSM >5.1.6.3 ... Start at the TSM Support page http://www-306.ibm.com/software/sysmgmt/products/support/IBMTivoliStorageMan ager.html and review Server Requirements. Richard Sims This e-mail and any files transmitted with it, are confidential and intended solely for the use of the addressee. The content of this e-mail may have been changed without the consent of the originator. The information supplied must be viewed in this context. If you have received this e-mail in error please notify our Helpdesk by telephone on +44 (0) 20-7444-8444. Any use, dissemination, forwarding, printing, or copying of this e-mail or its attachments is strictly prohibited.
auditing a Storagetek tape library with ACSLS
We have a Storagetek l5500 partitioned into 3 pieces. Tivoli has one piece with 400 slots. When I issue the following command AUDIT LIBR L5500 ( L5500 is the name of the library) It take a couple of hours for it to complete. Anyone know why it takes so long?? Thanx steve ** This email and any files transmitted with it are confidential and intended solely for the use of the individual or entity to whom they are addressed. If you have received this email in error please notify the system manager at postmaster at dor.state.ma.us. **
Re: auditing a Storagetek tape library with ACSLS
Add checkl=b to your command. If you don't, it will mount each tape in a drive to read the label. That adds lots of time! -Original Message- From: ADSM: Dist Stor Manager [mailto:[EMAIL PROTECTED] On Behalf Of Weinstein, Stephen Sent: Monday, January 31, 2005 12:58 PM To: ADSM-L@VM.MARIST.EDU Subject: auditing a Storagetek tape library with ACSLS We have a Storagetek l5500 partitioned into 3 pieces. Tivoli has one piece with 400 slots. When I issue the following command AUDIT LIBR L5500 ( L5500 is the name of the library) It take a couple of hours for it to complete. Anyone know why it takes so long?? Thanx steve ** This email and any files transmitted with it are confidential and intended solely for the use of the individual or entity to whom they are addressed. If you have received this email in error please notify the system manager at postmaster at dor.state.ma.us. **
Centera Boxes
I looked around in the archives here and found info on backing TO a Centera box but nothing about backing up the data FROM a Centera box. We are in the process of migrating assorted files on to our box and we need to see about backing up the data we move onto it. We are aware that the box itself has got all kinds of redundancy built in to protect the data that is stored on it but we are concerned about the box itself being lost due to fire, flood, locusts, or some IS guy going nuts with a shotgun. The vender said something about buying a mirror copy of what we already have but that's very expensive. Isn't there someway to backup the data on a Centera for purposes of disaster recovery? David Tyree Enterprise Backup Administrator South Georgia Medical Center 229.333.1155 Confidential Notice: This e-mail message, including any attachments, is for the sole use of the intended recipient(s) and may contain confidential and privileged information. Any unauthorized review, use, disclosure or distribution is prohibited. If you are not the intended recipient, please contact the sender by reply e-mail and destroy all copies of the original message.
Re: Centera Boxes
We're in the same boat ... The emc guys are telling us to use CASBAR (Content Addressable Storage Backup and Recovery). If I understand it all it does is extract the data from the centera and put on a file system of tivoli to pick up. Hope that helps. Aaron Durkee Infrastructure Analyst Networking and Technical Group Catholic Health Systems, Inc. [EMAIL PROTECTED] phone: (716) 862-1713 fax: (716) 862-1717 >>> [EMAIL PROTECTED] 01/31/05 01:44PM >>> I looked around in the archives here and found info on backing TO a Centera box but nothing about backing up the data FROM a Centera box. We are in the process of migrating assorted files on to our box and we need to see about backing up the data we move onto it. We are aware that the box itself has got all kinds of redundancy built in to protect the data that is stored on it but we are concerned about the box itself being lost due to fire, flood, locusts, or some IS guy going nuts with a shotgun. The vender said something about buying a mirror copy of what we already have but that's very expensive. Isn't there someway to backup the data on a Centera for purposes of disaster recovery? David Tyree Enterprise Backup Administrator South Georgia Medical Center 229.333.1155 Confidential Notice: This e-mail message, including any attachments, is for the sole use of the intended recipient(s) and may contain confidential and privileged information. Any unauthorized review, use, disclosure or distribution is prohibited. If you are not the intended recipient, please contact the sender by reply e-mail and destroy all copies of the original message.
EXPORT errors
I'm trying some machines to a different server. I've exported the client definitions for three large clients successfully. These will just cycle away in two months on their old box. I've tried exporting 5 different smaller clients, all with the same results: 01/31/05 12:06:26 ANR0919E Filespace name with length 0 was encountered for nodes CERIUM,STORM,STARBUCK,CALYPSO,MA,HONEYCOMB,PICASSO- .BSD.UCHICAGO.EDU,BUBBLES,MUTANTS,OUTGROUP.UCHICAGO.EDU,- CETUS,THERA,MHD2,BLAST,FLEETWOOD. (SESSION: 49773, PROCESS: 1410) 01/31/05 12:06:26 ANRD smxi.c(1165): ThreadId<155> Error 1 setting object names. Callchain of previous message: 0x10016138 outDiagf <- 0x104127D4 SmQryExportObjects <- 0x10433AC4 DoExportImportVerbs <- 0x104348C4 SmAdminSession <- 0x10404230 DoAdminGeneral <- 0x10407378 smExecuteSession <- 0x10417F3C XiSessionThread <- 0x10007AE8 StartThread <- 0xD004B410 _pthread_body <- (SESSION: 49773, PROCESS: 1410) The listed nodes with 0 length filespace names are in other domains, and are primarily IRIX machines with a few Linux86. Thay all have what seems to be healthy filespaces attached to them. If I try exporting the 5 nodes with filed=none, they export the definitions without problem. I suspect DB corruption, but at 200+ Gb for the DB, audit is out of the question. Both sending and receiving machines are AIX 5.2 and TSM 5.2.2.5. Until today and this domain, there hasn't been a hint of trouble with EXPORT NODE. Fred Johanson ITSM Administrator University of Chicago 773-702-8464
Re: EXPORT errors
Not sure this will yield anything of interest, but what does this show: select node_name, '"' || filespace_name || '"' from filespaces where node_name='CERIUM' (or pick any other node name that is having this difficulty). 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] 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 01/31/2005 14:42:26: > I'm trying some machines to a different server. I've exported the client > definitions for three large clients successfully. These will just cycle > away in two months on their old box. I've tried exporting 5 different > smaller clients, all with the same results: > > > >01/31/05 12:06:26 ANR0919E Filespace name with length 0 was > >encountered for > > nodes > > CERIUM,STORM,STARBUCK,CALYPSO,MA,HONEYCOMB,PICASSO- > > > >.BSD.UCHICAGO.EDU,BUBBLES,MUTANTS,OUTGROUP.UCHICAGO.EDU,- > > CETUS,THERA,MHD2,BLAST,FLEETWOOD. (SESSION: > > 49773, > > PROCESS: > > 1410) > >01/31/05 12:06:26 ANRD smxi.c(1165): ThreadId<155> Error 1 > >setting > > object names. Callchain of previous message: > > 0x10016138 > > outDiagf <- 0x104127D4 SmQryExportObjects <- > > 0x10433AC4 > > DoExportImportVerbs <- 0x104348C4 > > SmAdminSession <- > > 0x10404230 DoAdminGeneral <- 0x10407378 > > smExecuteSession > > <- 0x10417F3C XiSessionThread <- 0x10007AE8 > > StartThread > > <- 0xD004B410 _pthread_body <- (SESSION: > > 49773, PROCESS: > > 1410) > > The listed nodes with 0 length filespace names are in other domains, and > are primarily IRIX machines with a few Linux86. Thay all have what seems > to be healthy filespaces attached to them. If I try exporting the 5 nodes > with filed=none, they export the definitions without problem. I suspect DB > corruption, but at 200+ Gb for the DB, audit is out of the question. > > Both sending and receiving machines are AIX 5.2 and TSM 5.2.2.5. Until > today and this domain, there hasn't been a hint of trouble with EXPORT NODE. > > > > > > > Fred Johanson > ITSM Administrator > University of Chicago > 773-702-8464
Re: EXPORT errors
Each client on the list has a filespace with an fsid, but no name. At 02:52 PM 1/31/2005 -0700, you wrote: Not sure this will yield anything of interest, but what does this show: select node_name, '"' || filespace_name || '"' from filespaces where node_name='CERIUM' (or pick any other node name that is having this difficulty). 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] 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 01/31/2005 14:42:26: > I'm trying some machines to a different server. I've exported the client > definitions for three large clients successfully. These will just cycle > away in two months on their old box. I've tried exporting 5 different > smaller clients, all with the same results: > > > >01/31/05 12:06:26 ANR0919E Filespace name with length 0 was > >encountered for > > nodes > > CERIUM,STORM,STARBUCK,CALYPSO,MA,HONEYCOMB,PICASSO- > > > >.BSD.UCHICAGO.EDU,BUBBLES,MUTANTS,OUTGROUP.UCHICAGO.EDU,- > > CETUS,THERA,MHD2,BLAST,FLEETWOOD. (SESSION: > > 49773, > > PROCESS: > > 1410) > >01/31/05 12:06:26 ANRD smxi.c(1165): ThreadId<155> Error 1 > >setting > > object names. Callchain of previous message: > > 0x10016138 > > outDiagf <- 0x104127D4 SmQryExportObjects <- > > 0x10433AC4 > > DoExportImportVerbs <- 0x104348C4 > > SmAdminSession <- > > 0x10404230 DoAdminGeneral <- 0x10407378 > > smExecuteSession > > <- 0x10417F3C XiSessionThread <- 0x10007AE8 > > StartThread > > <- 0xD004B410 _pthread_body <- (SESSION: > > 49773, PROCESS: > > 1410) > > The listed nodes with 0 length filespace names are in other domains, and > are primarily IRIX machines with a few Linux86. Thay all have what seems > to be healthy filespaces attached to them. If I try exporting the 5 nodes > with filed=none, they export the definitions without problem. I suspect DB > corruption, but at 200+ Gb for the DB, audit is out of the question. > > Both sending and receiving machines are AIX 5.2 and TSM 5.2.2.5. Until > today and this domain, there hasn't been a hint of trouble with EXPORT NODE. > > > > > > > Fred Johanson > ITSM Administrator > University of Chicago > 773-702-8464 Fred Johanson ITSM Administrator University of Chicago 773-702-8464
Re: EXPORT errors
OK, that is in agreement with what the ANR0919E message is telling you, though I have no idea how you could end up with file spaces with no names. Can you paste in the exact output from the SELECT statement, as well as output from QUERY FILESPACE F=D, for one or two of these problem nodes? 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] 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 01/31/2005 15:03:42: > Each client on the list has a filespace with an fsid, but no name. > > > At 02:52 PM 1/31/2005 -0700, you wrote: > >Not sure this will yield anything of interest, but what does this show: > > > >select node_name, '"' || filespace_name || '"' from filespaces where > >node_name='CERIUM' > > > >(or pick any other node name that is having this difficulty). > > > >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] > > > >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 01/31/2005 > >14:42:26: > > > > > I'm trying some machines to a different server. I've exported the > >client > > > definitions for three large clients successfully. These will just cycle > > > away in two months on their old box. I've tried exporting 5 different > > > smaller clients, all with the same results: > > > > > > > > > >01/31/05 12:06:26 ANR0919E Filespace name with length 0 was > > > >encountered for > > > > nodes > > > > CERIUM,STORM,STARBUCK,CALYPSO,MA,HONEYCOMB,PICASSO- > > > > > > > >.BSD.UCHICAGO.EDU,BUBBLES,MUTANTS,OUTGROUP.UCHICAGO.EDU,- > > > > CETUS,THERA,MHD2,BLAST,FLEETWOOD. (SESSION: > > > > 49773, > > > > PROCESS: > > > > 1410) > > > >01/31/05 12:06:26 ANRD smxi.c(1165): ThreadId<155> Error 1 > > > >setting > > > > object names. Callchain of previous message: > > > > 0x10016138 > > > > outDiagf <- 0x104127D4 SmQryExportObjects <- > > > > 0x10433AC4 > > > > DoExportImportVerbs <- 0x104348C4 > > > > SmAdminSession <- > > > > 0x10404230 DoAdminGeneral <- 0x10407378 > > > > smExecuteSession > > > > <- 0x10417F3C XiSessionThread <- 0x10007AE8 > > > > StartThread > > > > <- 0xD004B410 _pthread_body <- (SESSION: > > > > 49773, PROCESS: > > > > 1410) > > > > > > The listed nodes with 0 length filespace names are in other domains, and > > > are primarily IRIX machines with a few Linux86. Thay all have what > >seems > > > to be healthy filespaces attached to them. If I try exporting the 5 > >nodes > > > with filed=none, they export the definitions without problem. I suspect > >DB > > > corruption, but at 200+ Gb for the DB, audit is out of the question. > > > > > > Both sending and receiving machines are AIX 5.2 and TSM 5.2.2.5. Until > > > today and this domain, there hasn't been a hint of trouble with EXPORT > >NODE. > > > > > > > > > > > > > > > > > > > Fred Johanson > > > ITSM Administrator > > > University of Chicago > > > 773-702-8464 > > Fred Johanson > ITSM Administrator > University of Chicago > 773-702-8464
Re: EXPORT errors
One other question: can the actual client nodes connect to the TSM server and restore data from these file spaces? Andy Raibeck IBM Software Group Tivoli Storage Manager Client Development Internal Notes e-mail: Andrew Raibeck/Tucson/[EMAIL PROTECTED] Internet e-mail: [EMAIL PROTECTED] 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 01/31/2005 15:15:08: > OK, that is in agreement with what the ANR0919E message is telling you, > though I have no idea how you could end up with file spaces with no names. > > Can you paste in the exact output from the SELECT statement, as well as > output from QUERY FILESPACE F=D, for one or two of these problem nodes? > > 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] > > 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 01/31/2005 > 15:03:42: > > > Each client on the list has a filespace with an fsid, but no name. > > > > > > At 02:52 PM 1/31/2005 -0700, you wrote: > > >Not sure this will yield anything of interest, but what does this show: > > > > > >select node_name, '"' || filespace_name || '"' from filespaces where > > >node_name='CERIUM' > > > > > >(or pick any other node name that is having this difficulty). > > > > > >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] > > > > > >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 01/31/2005 > > >14:42:26: > > > > > > > I'm trying some machines to a different server. I've exported the > > >client > > > > definitions for three large clients successfully. These will just > cycle > > > > away in two months on their old box. I've tried exporting 5 > different > > > > smaller clients, all with the same results: > > > > > > > > > > > > >01/31/05 12:06:26 ANR0919E Filespace name with length 0 was > > > > >encountered for > > > > > nodes > > > > > CERIUM,STORM,STARBUCK,CALYPSO,MA,HONEYCOMB,PICASSO- > > > > > > > > > >.BSD.UCHICAGO.EDU,BUBBLES,MUTANTS,OUTGROUP.UCHICAGO.EDU,- > > > > > CETUS,THERA,MHD2,BLAST,FLEETWOOD. > (SESSION: > > > > > 49773, > > > > > PROCESS: > > > > > 1410) > > > > >01/31/05 12:06:26 ANRD smxi.c(1165): ThreadId<155> Error > 1 > > > > >setting > > > > > object names. Callchain of previous > message: > > > > > 0x10016138 > > > > > outDiagf <- 0x104127D4 > SmQryExportObjects <- > > > > > 0x10433AC4 > > > > > DoExportImportVerbs <- 0x104348C4 > > > > > SmAdminSession <- > > > > > 0x10404230 DoAdminGeneral <- 0x10407378 > > > > > smExecuteSession > > > > > <- 0x10417F3C XiSessionThread <- > 0x10007AE8 > > > > > StartThread > > > > > <- 0xD004B410 _pthread_body <- (SESSION: > > > > > 49773, PROCESS: > > > > > 1410) > > > > > > > > The listed nodes with 0 length filespace names are in other domains, > and > > > > are primarily IRIX machines with a few Linux86. Thay all have what > > >seems > > > > to be healthy filespaces attached to them. If I try exporting the 5 > > >nodes > > > > with filed=none, they export the definitions without problem. I > suspect > > >DB > > > > corruption, but at 200+ Gb for the DB, audit is out of the question. > > > > > > > > Both sending and receiving machines are AIX 5.2 and TSM 5.2.2.5. > Until > > > > today and this domain, there hasn't been a hint of trouble with > EXPORT > > >NODE. > > > > > > > > > > > > > > > > > > > > > > > > > Fred Johanson > > > > ITSM Administrator > > > > University of Chicago > > > > 773-702-8464 > > > > Fred Johanson > > ITSM Administrator > > University of Chicago > > 773-702-8464
Export error, cont.
Andy, Here's the first one I tried. Q OCC only shows FSID 1 and 2. Unnamed[1] -- CERIUM CERIUM/ CERIUM/nfs/cerium- /c1 tsm: ADSM>q fi cerium Node NameFilespaceFSID Platform Filespace Is Files- CapacityPct Name Type pace (MB) Util Unicode? --- --- - - - CERIUM / 1 IRIX XFS No 4,153.8 59.9 CERIUM /nfs/ceriu- 2 IRIX XFS No 17,360.5 32.2 m/c1 CERIUM 3 IRIX XFS No 4,153.8 61.3 Fred Johanson ITSM Administrator University of Chicago 773-702-8464
Re: EXPORT errors
I'll have to find an owner and try. Tomorrow, since it's leaving time here. What do you think would happen if I tried to delete all the phantoms from my list a ran the exp node again? At 03:18 PM 1/31/2005 -0700, you wrote: One other question: can the actual client nodes connect to the TSM server and restore data from these file spaces? Andy Raibeck IBM Software Group Tivoli Storage Manager Client Development Internal Notes e-mail: Andrew Raibeck/Tucson/[EMAIL PROTECTED] Internet e-mail: [EMAIL PROTECTED] 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 01/31/2005 15:15:08: > OK, that is in agreement with what the ANR0919E message is telling you, > though I have no idea how you could end up with file spaces with no names. > > Can you paste in the exact output from the SELECT statement, as well as > output from QUERY FILESPACE F=D, for one or two of these problem nodes? > > 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] > > 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 01/31/2005 > 15:03:42: > > > Each client on the list has a filespace with an fsid, but no name. > > > > > > At 02:52 PM 1/31/2005 -0700, you wrote: > > >Not sure this will yield anything of interest, but what does this show: > > > > > >select node_name, '"' || filespace_name || '"' from filespaces where > > >node_name='CERIUM' > > > > > >(or pick any other node name that is having this difficulty). > > > > > >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] > > > > > >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 01/31/2005 > > >14:42:26: > > > > > > > I'm trying some machines to a different server. I've exported the > > >client > > > > definitions for three large clients successfully. These will just > cycle > > > > away in two months on their old box. I've tried exporting 5 > different > > > > smaller clients, all with the same results: > > > > > > > > > > > > >01/31/05 12:06:26 ANR0919E Filespace name with length 0 was > > > > >encountered for > > > > > nodes > > > > > CERIUM,STORM,STARBUCK,CALYPSO,MA,HONEYCOMB,PICASSO- > > > > > > > > > >.BSD.UCHICAGO.EDU,BUBBLES,MUTANTS,OUTGROUP.UCHICAGO.EDU,- > > > > > CETUS,THERA,MHD2,BLAST,FLEETWOOD. > (SESSION: > > > > > 49773, > > > > > PROCESS: > > > > > 1410) > > > > >01/31/05 12:06:26 ANRD smxi.c(1165): ThreadId<155> Error > 1 > > > > >setting > > > > > object names. Callchain of previous > message: > > > > > 0x10016138 > > > > > outDiagf <- 0x104127D4 > SmQryExportObjects <- > > > > > 0x10433AC4 > > > > > DoExportImportVerbs <- 0x104348C4 > > > > > SmAdminSession <- > > > > > 0x10404230 DoAdminGeneral <- 0x10407378 > > > > > smExecuteSession > > > > > <- 0x10417F3C XiSessionThread <- > 0x10007AE8 > > > > > StartThread > > > > > <- 0xD004B410 _pthread_body <- (SESSION: > > > > > 49773, PROCESS: > > > > > 1410) > > > > > > > > The listed nodes with 0 length filespace names are in other domains, > and > > > > are primarily IRIX machines with a few Linux86. Thay all have what > > >seems > > > > to be healthy filespaces attached to them. If I try exporting the 5 > > >nodes > > > > with filed=none, they export the definitions without problem. I > suspect > > >DB > > > > corruption, but at 200+ Gb for the DB, audit is out of the question. > > > > > > > > Both sending and receiving machines are AIX 5.2 and TSM 5.2.2.5. > Until > > > > today and this domain, there hasn't been a hint of trouble with > EXPORT > > >NODE. > > > > > > > > > > > > > > > > > > > > > > > > > Fred Johanson > > > > ITSM Administrator > > > > University of Chicago > > > > 773-702-8464 > > > > Fred Johanson > > ITSM Administrator > > University of Chicago > > 773-702-8464 Fred Johanson ITSM Administrator University of Chicago 773-702-8464
Re: Export error, cont.
OK... actually the SELECT statement I was looking for: select node_name, '"' || filespace_name || '"' from filespaces where node_name='CERIUM' was intended to provide the node name in one output column and the filespace name, in double quotes, in the second output column, so we could clearly see the empty name. Also, would be helpful to see the detailed output of Q FI. Do you have any idea what these blank-named file spaces might represent (i.e. what might have been stored from the original client machines)? In the output you show here, the capacity and utilization are awful close to those of the root (/) file system. I think that at this point, you should open a call with support. I can't explain the blank file space names, and obviously they are causing the export to fail. 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] 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 01/31/2005 15:22:00: > Andy, > > Here's the first one I tried. > > Q OCC only shows FSID 1 and 2. > > > > > >Unnamed[1] > >-- > >CERIUM > >CERIUM/ > >CERIUM/nfs/cerium- > > /c1 > > > >tsm: ADSM>q fi cerium > > > >Node NameFilespaceFSID Platform Filespace Is > >Files- CapacityPct > > Name Type pace > > (MB) Util > > Unicode? > > > >--- --- - - > > - > >CERIUM / 1 IRIX XFS No > >4,153.8 59.9 > >CERIUM /nfs/ceriu- 2 IRIX XFS No > >17,360.5 32.2 > > m/c1 > > > >CERIUM 3 IRIX XFS No > >4,153.8 61.3 > > Fred Johanson > ITSM Administrator > University of Chicago > 773-702-8464
TSM Restore on Win2K machine
I have a question on TSM restore on a Windows 2000(SP3) machine running TSM 5.2.2. I tried to test to see if I could successfully restore a machine after a backup. What I did was before backing up I ran NTBackup to backup the SYSTEM OBJECT then ran a TSM backup on the drives. I then rebuilt the machine with just win2K(without SP3) and installed TSM. I restored the 2 data drives, when it asked to reboot I said no, then I ran NTBackup to restore the SYSTEM OBJECT and rebooted. The problem is when I rebooted and looked in the event log there were a few services that did not run and there were messages during the restore that some of the files were protected. Also at the end of the restore it said to install Windows SP3 because there were unknown files and it may not run correctly. I did not install the CD(SP3) thinking the TSM restore would restore these files. Is there a step I'm doing wrong or missing? The TSM server is AIX 5.2 with TSM 5.2.2 Have a Great Day, Eric Jones > * [EMAIL PROTECTED] > *: 607-751-4133 Cell : 607-972-7621
Re: EXPORT errors
I would hesitate to delete the "phantom" file spaces (if TSM lets you) unless you know they really aren't necessary. I agree, if Q OCC shows nothing, then the risk is probably minimal; just the same, I wouldn't be comfortable suggesting you delete them until you've had a chance to let the support team have a look-see. 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] 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 01/31/2005 15:31:27: > I'll have to find an owner and try. Tomorrow, since it's leaving time > here. What do you think would happen if I tried to delete all the phantoms > from my list a ran the exp node again? > > > At 03:18 PM 1/31/2005 -0700, you wrote: > >One other question: can the actual client nodes connect to the TSM server > >and restore data from these file spaces? > > > >Andy Raibeck > >IBM Software Group > >Tivoli Storage Manager Client Development > >Internal Notes e-mail: Andrew Raibeck/Tucson/[EMAIL PROTECTED] > >Internet e-mail: [EMAIL PROTECTED] > > > >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 01/31/2005 > >15:15:08: > > > > > OK, that is in agreement with what the ANR0919E message is telling you, > > > though I have no idea how you could end up with file spaces with no > >names. > > > > > > Can you paste in the exact output from the SELECT statement, as well as > > > output from QUERY FILESPACE F=D, for one or two of these problem nodes? > > > > > > 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] > > > > > > 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 01/31/2005 > > > 15:03:42: > > > > > > > Each client on the list has a filespace with an fsid, but no name. > > > > > > > > > > > > At 02:52 PM 1/31/2005 -0700, you wrote: > > > > >Not sure this will yield anything of interest, but what does this > >show: > > > > > > > > > >select node_name, '"' || filespace_name || '"' from filespaces where > > > > >node_name='CERIUM' > > > > > > > > > >(or pick any other node name that is having this difficulty). > > > > > > > > > >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] > > > > > > > > > >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 01/31/2005 > > > > >14:42:26: > > > > > > > > > > > I'm trying some machines to a different server. I've exported the > > > > >client > > > > > > definitions for three large clients successfully. These will just > > > cycle > > > > > > away in two months on their old box. I've tried exporting 5 > > > different > > > > > > smaller clients, all with the same results: > > > > > > > > > > > > > > > > > > >01/31/05 12:06:26 ANR0919E Filespace name with length 0 > >was > > > > > > >encountered for > > > > > > > nodes > > > > > > > CERIUM,STORM,STARBUCK,CALYPSO,MA,HONEYCOMB,PICASSO- > > > > > > > > > > > > > >.BSD.UCHICAGO.EDU,BUBBLES,MUTANTS,OUTGROUP.UCHICAGO.EDU,- > > > > > > > CETUS,THERA,MHD2,BLAST,FLEETWOOD. > > > (SESSION: > > > > > > > 49773, > > > > > > > PROCESS: > > > > > > > 1410) > > > > > > >01/31/05 12:06:26 ANRD smxi.c(1165): ThreadId<155> > >Error > > > 1 > > > > > > >setting > > > > > > > object names. Callchain of previous > > > message: > > > > > > > 0x10016138 > > > > > > > outDiagf <- 0x104127D4 > > > SmQryExportObjects <- > > > > > > > 0x10433AC4 > > > > > > > DoExportImportVerbs <- 0x104348C4 > > > > > > > SmAdminSession <- > > > > > > > 0x10404230 DoAdminGeneral <- > >0x10407378 > > > > > > > smExecuteSession > > > > > > > <- 0x10417F3C XiSessionThread <- > > > 0x10007AE8 > > > > > > > StartThread > > > > > > > <- 0xD004B410 _pthread_body <- > >(SESSION: > > > > > > > 49773, PROCESS: > > > > > > > 1410) > > > > > > > > > > > > The listed nodes with 0 length filespace names are in other > >domai
dsmc select, etc.
Many days ago I asked for (and received, thank you) help about
'dsmc select' and 'dsmc expire' with the '-filelist' option. Here
there has been installed a 'Vault' like a HMC that moves files and
email messages that have not been used in some time out of the file
server and exchange server (can you tell we're using Microsoft)
to a different location. The 'Vault' server leaves behind 'stub'
files and the hooks in the OS allow for users to open the stub and
receive the file as if the file were still local to the file server.
This behavior is causing the TSM server some problems. The OS
hooks do set the 'O' or 'Offline' flag when a file is moved and a
stub remains.
Attached (if the list server lets me do attachments) is the program
I mentioned that I would pass on once I'd finished writing it. The
name is 'dsmpmj' for 'DSM - Poor Man's Journal' since the program
also does the journalling feature of the B/A client.
The program reads yesterday's file, scans the local filesystem,
and then compares the two lists. Anything that existed yesterday,
but is not there today goes into the EXPIRELIST. Anything that
has a timestamp newer than the timestamp of the YESTERDAY file goes
into BACKUPLIST. If any part of the lines in EXCLUDELIST match the
local files, the files are excluded. That means if BACKUPLIST has
a line '.sys' this will match the file 'c:\user\intricate.system',
the latter file not making it to BACKUPLIST.
The way to run this is:
dsmpmj -y YESTERDAY -b BACKUPLIST -e EXPIRELIST -x EXCLUDELIST
There is also a -r REFERENCEFILE that can be used the first time
the program is called to give dsmpmj a place to start for it's
backup timestamp (i.e. four days ago).
Once you have run dsmpmj feed the two generated files to dsmc:
c:\> dsmpmj -y yesterday -b backup -e expire -x excludes
c:\> dsmc expire -filelist=expire
c:\> dsmc sel -filelist=backup
That's about it. I hope it works for other people.
Mike
static char *rcsid = "$Id$";
/*
* $Log$
*/
/* generate a list of all files on the server
* accept a list of all files in the TSM library
* compare the two lists of files
* output the differences... files on the server not in the library
*/
#include
#include
#include
#include
#include
#include
#include
typedef struct _file {
FILETIME mtime;
unsigned int size;
unsigned int flags;
char *name;
} File;
/* global variables */
int verbose = 0;
/* logfile */
static FILE *logfp;
static char *logfn;
/* simple error routing */
void err(char *msg)
{
(void) fprintf(stderr, "dsmpmj: %s\n", msg);
if(logfp) {
(void) fprintf(logfp, "dsmpmj: %s\n", msg);
(void) fclose(logfp);
}
(void) exit(1);
}
/* warning messages */
void warn(char *msg)
{
time_t t;
char buf[256];
(void) time(&t);
(void) strcpy(buf, ctime(&t));
if(strchr(buf, '\n')) {
*(strchr(buf, '\n')) = '\0';
}
(void) fprintf(stderr, "%s: dsmpmj: %s\n", buf, msg);
if(logfp) {
(void) fprintf(logfp, "%s: dsmpmj: %s\n", buf, msg);
}
}
/* comparison function for qsort */
int filesCompare(const void *a, const void *b)
{
File *af = (File *) a;
File *bf = (File *) b;
return strcmp(af->name, bf->name);
}
/* comparison function for bsearch */
int keyCompare(const void *a, const void *b)
{
File *bf = (File *) b;
return strcmp((char *) a, bf->name);
}
/* change text to lower case */
void toLower(char *s)
{
while(s && *s) {
*s = tolower(*s);
s++;
}
}
/* load the file previos day's run list of files */
File *loadFiles(char *fn, int *n, int missingok)
{
int i, nfiles;
FILE *fp;
File *files;
char buf[1024], *p;
if((fp = fopen(fn, "r")) == (FILE *) NULL) {
if(missingok) {
*n = 0;
return (File*) NULL;
}
(void) sprintf(buf, "unable to read file '%s'", fn);
err(buf);
}
/* how many files found? */
if(verbose) {
warn("counting output from filesystem scans");
}
(void) fseek(fp, 0, SEEK_SET); /* rewind to beginning of file
*/
nfiles = 0;
while(fgets(buf, sizeof(buf) - 1, fp)) {/* count number of
lines/files in fn */
nfiles++;
}
if(verbose) {
(void) sprintf(buf, "counted a total of %d filenames", nfiles);
warn(buf);
}
/* allocate memory for the file data */
if((files = (File *) malloc(sizeof(File) * nfiles)) == (File *) NULL) {
err("unable to allocate memory for local files");
}
if(verbose) {
(void) sprintf(buf, "allocate %d bytes", sizeof(File) * nfiles);
warn(buf);
}
/* load t
Re: Export error, cont.
>OK, that is in agreement with what the ANR0919E message is telling you, >though I have no idea how you could end up with file spaces with no names. ... May be circumstances as described in APAR IC32186 and Technote 1194088. As Andy says, we need the Format=Detailed output of Query Filespace, which will report the filespace name in hex, so we can determine what's actually in there. (Beware non-printing, space-occupying characters.) Richard Sims
Re: NEXTSTGPOOL with MAXSIZE does not work with MOVE DATA
I see nobody else has jumped into this, so I will. I think IBM Support is right. MOVE DATA and Reclamation are fundamentally different from Migration and Client Backup. They're potentially going in the opposite direction - either sideways in the Storage Heirarchy, or downwards - whereas Migration and Client Backup are both going upwards. The manual for MOVE DATA says: STGpool Specifies the primary storage pool to which you want to move files (the target storage pool). This parameter is optional and applies only to moving data from primary storage pool volumes. If you do not specify a value for this parameter, files are moved to other volumes within the same storage pool. This is quite clear. The target storage pool is THE storage pool it will move the files to. It does not say that it will move it to that point in the Storage Heirarchy, including anywhere uphill it has to due to MAXSIZE, but it says to THAT pool. If the file doesn't fit due to MAXSIZE, then the operation should fail. This does leave you with how to solve your problem. I think doing MOVE DATA to the Disk Storage Pool is about the best way. That way, you are getting things into the Storage Heirarchy in the way intended, letting Migration sort the files out by size, efficiently filling volumes higher up in the heirarchy. This was the basic philosophy driving the original design of the system, and it's at this basic design philosophy level that I have to agree with IBM Support on this point. Throw some extra temporary disk space at that Disk storage pool and make it huge. Or perhaps define a separate temporary disk storage pool just for this project, to reduce the performance penalty for your regular backups. It could be a FILE disk storage pool, using all your favorite RAID tricks to benefit large sequential files such as RAID0 Striping, and it would fly, both on the way in and the way back out. Who knows - you could be collocating. That could make a real mess of this. Roger Deschner University of Illinois at Chicago [EMAIL PROTECTED] ==I have not lost my mind -- it is backed up on tape somewhere.= On Mon, 31 Jan 2005, Rushforth, Tim wrote: >We have a Storage Pool hierarchy something like: > > > >DISK Stgpool (noliimt) --> File Stgpool (2GB limit) --> Tape Stgpool >(nolimit) > > > >Clients backup to DISK, files smaller than 2GB migrate to "File" >stgpool, everything else migrates to Tape. > > > >We've changed the MAXSIZE to 2GB and wanted to move all files from Tape >that are smaller than 2GB to the File Stgpool. This can be done by >moving everything to DISK and let migration handle it. > > > >I thought it would be more efficient to just move this data to the File >stgpool but ran into the following error: > >ANR1150W Move data process terminated for volume VOLNAME - > unable to move file to target storage pool due to > exclusion by size. > > > > > >So I thought it was weird that Migration will honor the Nextstgpool when >a maxsize is set but MOVE DATA does not. > > > >(Also client backups/archives will honor the nextstgpool). > > > >Along with MOVE DATA, MOVE NODEDATA and RECLAMATION also do not work in >this scenario. > > > >I called Tivoli Support and they said working as designed and are going >to update the documentation to state this. > > > >My questions are: > > > >- does anyone else think this is working a bit weird (ie. >Migration and client sessions work fine when writing to a stgpool with a >maxsize and nextstgpool but move data, move nodedata and reclamation do >not). > >- Any other (efficient) way of doing what I want - move the >small files from tape to the file pool > > > >Thanks, > > > >Tim Rushforth > >City of Winnipeg >
