Test Restoration failed

Hello I was trying to restore a level 0 from our dev server on to my test server both are IDS 11.10, but the dev server had 4 data chunks and i build the test server with only 3 chunks, rest all the chunks and their path are similar to that of dev server. ( I made a mistake my not adding the 4th chunk and makinig every thing identical which should have worked perfectly ) Running "ontape -r" I got the error as oninit: Cannot open chunk '/informix/db/datadbs2_04'. errno = 2 Physical restore failed - Cannot Open Primary Chunk '/informix/db/datadbs2_04'. 11:26:16 ipcshm buf_wait semop errno=36 And from fast recovery the database went down.tried to take it online with "oninit" ( i did not do any onmode -ky )and go the error as Initializing log/checkpoint information...succeeded Initializing dbspaces...succeeded Opening primary chunks...Bad Primary Chunk '/informix/db/datadbs2_04'. succeeded Opening mirror chunks...succeeded Validating chunks...succeeded Initialize Async Log Flusher...succeeded Forking btree cleaner...succeeded Initializing DBSPACETEMP list...succeeded Checking database partition index...FAILED oninit: Fatal error in shared memory initialization in the path /informix/db/ is see a 0 byte file created "datadbs2_04" I tired onspaces -s datadbs2 -p /informix/db/datadbs2_04 -o 0 -D -ydid not worked and got shared mem not initialized. What possibly can i do here.Or should i have to reinitialize the test instance again? Not sure but i guess there is some option which allows me to do such kind of restoration.

Hi, I think best is to create that fourth chunk (or make sure that it has correct ownership and file access rights by now) and then perform the restore again from scratch. You have to restore the root dbspace and all critical dbspaces (containing the physical log file and/or the logical log files). Other dbspaces (containing 'only' user data) need not be restored at time of the level 0 cold restore. They can be restored later by doing a warm restore (that however requires that you have backed up logical logs and can restore them). Saying "you have to restore the root dbspace ..." means that all chunks that belong to root dbspace and critical dbspaces must be restored. As you are talking about "chunks" I assume, that this was a chunk that belonged to the root dbspace - as such it must be restored with the level 0 cold restore. Regards, Martin -- Martin Fuerderer IBM Informix Development Munich, Germany Information Management IBM Deutschland Research & Development GmbH Chairman of the Supervisory Board: Martin Jetter Board of Management: Erich Baier Corporate Seat: Boeblingen, Germany Reg.-Gericht: Amtsgericht Stuttgart, HRB 243294 ids-bounces@iiug.org wrote on 07.08.2008 13:38:02: > Hello > > I was trying to restore a level 0 from our dev server on to my test server > both are IDS 11.10, but the dev server had 4 data chunks and i buildthe test > server with only 3 chunks, rest all the chunks and their path are similar to > that of dev server. ( I made a mistake my not adding the 4th chunk > and makinig > every thing identical which should have worked perfectly ) > > Running "ontape -r" I got the error as > oninit: Cannot open chunk '/informix/db/datadbs2_04'. errno > = 2 > Physical restore failed - Cannot Open Primary Chunk > '/informix/db/datadbs2_04'. > 11:26:16 ipcshm buf_wait semop errno=36 > > And from fast recovery the database went down.tried to take it online with > "oninit" ( i did not do any onmode -ky )and go the error as > > Initializing log/checkpoint information...succeeded > Initializing dbspaces...succeeded > Opening primary chunks...Bad Primary Chunk '/informix/db/datadbs2_04'. > succeeded > Opening mirror chunks...succeeded > Validating chunks...succeeded > Initialize Async Log Flusher...succeeded > Forking btree cleaner...succeeded > Initializing DBSPACETEMP list...succeeded > Checking database partition index...FAILED > oninit: Fatal error in shared memory initialization > > in the path /informix/db/ is see a 0 byte file created "datadbs2_04" > > I tired > onspaces -s datadbs2 -p /informix/db/datadbs2_04 -o 0 -D -y> did not worked and got shared mem not initialized. > > What possibly can i do here.Or should i have to reinitialize the > test instance > again? > > Not sure but i guess there is some option which allows me to do such kind of > restoration. > > > ******************************************************************************* > Forum Note: Use "Reply" to post a response in the discussion forum. >

My dev server has rootdbs,physdbs,logdbs,datadbs2_01,datadbs2_02,datadbs2_03,datadbs2_04,tempdbs_0 1 and tempdbs_02. The newly insalled test server had rootdbs,physdbs,logdbs,datadbs2_01,datadbs2_02,datadbs2_03,tempdbs_01 and tempdbs_02. so databds_04 was missing(not created)on test instance. I had used a Level 0 backup of dev instance for the cold restoration of test instance. How do i create that fourth chunk when my test instance in not coming online. Is there some thing i missing out here?? >Hi, >I think best is to create that fourth chunk (or make sure that it >has correct ownership and file access rights by now) and then >perform the restore again from scratch. >You have to restore the root dbspace and all critical dbspaces >(containing the physical log file and/or the logical log files). >Other dbspaces (containing 'only' user data) need not be >restored at time of the level 0 cold restore. They can be >restored later by doing a warm restore (that however >requires that you have backed up logical logs and can >restore them). >Saying "you have to restore the root dbspace ..." means >that all chunks that belong to root dbspace and critical >dbspaces must be restored. As you are talking about >"chunks" I assume, that this was a chunk that belonged >to the root dbspace - as such it must be restored with >the level 0 cold restore. >Regards, >Martin

Hi, I think what martin was saying, is that you need to create the missing chunk as an empty data file with the correct permissions and ownership, just as you prepared the datadbs2_03 file before you ran the onspaces command. Then re-run the ontape -r -----Original Message----- From: ids-bounces@iiug.org [mailto:ids-bounces@iiug.org] On Behalf Of VICKY H Sent: 07 August 2008 02:32 PM To: ids@iiug.org Subject: Re: Test Restoration failed [13014] My dev server has rootdbs,physdbs,logdbs,datadbs2_01,datadbs2_02,datadbs2_03,datadbs2_04,tempd bs_01 and tempdbs_02. The newly insalled test server had rootdbs,physdbs,logdbs,datadbs2_01,datadbs2_02,datadbs2_03,tempdbs_01 and tempdbs_02. so databds_04 was missing(not created)on test instance. I had used a Level 0 backup of dev instance for the cold restoration of test instance. How do i create that fourth chunk when my test instance in not coming online. Is there some thing i missing out here?? >Hi, >I think best is to create that fourth chunk (or make sure that it has >correct ownership and file access rights by now) and then perform the >restore again from scratch. >You have to restore the root dbspace and all critical dbspaces >(containing the physical log file and/or the logical log files). >Other dbspaces (containing 'only' user data) need not be restored at >time of the level 0 cold restore. They can be restored later by doing a >warm restore (that however requires that you have backed up logical >logs and can restore them). >Saying "you have to restore the root dbspace ..." means that all chunks >that belong to root dbspace and critical dbspaces must be restored. As >you are talking about "chunks" I assume, that this was a chunk that >belonged to the root dbspace - as such it must be restored with the >level 0 cold restore. >Regards, >Martin **************************************************************************** *** Forum Note: Use "Reply" to post a response in the discussion forum.

Hi, yes, Mark, that's what I was trying to say. However, IDS can start and be on-line while a user data dbspace is missing or corrupt. There is an onconfig file parameter named ONDBSPACEDOWN. IDS will continue when this is set to 0 (zero), it will abort when this is set to 1 and it will wait (hang) when this is set to 2. So maybe on the system in question this is set to 1 and that's why IDS aborts? TIA, Martin -- Martin Fuerderer IBM Informix Development Munich, Germany Information Management IBM Deutschland Research & Development GmbH Chairman of the Supervisory Board: Martin Jetter Board of Management: Erich Baier Corporate Seat: Boeblingen, Germany Reg.-Gericht: Amtsgericht Stuttgart, HRB 243294 "Mark Tyrer" <mark.tyrer@rtt.co.za> Sent by: ids-bounces@iiug.org 07.08.2008 14:42 Please respond to ids@iiug.org To ids@iiug.org cc Subject RE: Test Restoration failed [13015] Hi, I think what martin was saying, is that you need to create the missing chunk as an empty data file with the correct permissions and ownership, just as you prepared the datadbs2_03 file before you ran the onspaces command. Then re-run the ontape -r -----Original Message----- From: ids-bounces@iiug.org [mailto:ids-bounces@iiug.org] On Behalf Of VICKY H Sent: 07 August 2008 02:32 PM To: ids@iiug.org Subject: Re: Test Restoration failed [13014] My dev server has rootdbs,physdbs,logdbs,datadbs2_01,datadbs2_02,datadbs2_03,datadbs2_04,tempd bs_01 and tempdbs_02. The newly insalled test server had rootdbs,physdbs,logdbs,datadbs2_01,datadbs2_02,datadbs2_03,tempdbs_01 and tempdbs_02. so databds_04 was missing(not created)on test instance. I had used a Level 0 backup of dev instance for the cold restoration of test instance. How do i create that fourth chunk when my test instance in not coming online. Is there some thing i missing out here?? >Hi, >I think best is to create that fourth chunk (or make sure that it has >correct ownership and file access rights by now) and then perform the >restore again from scratch. >You have to restore the root dbspace and all critical dbspaces >(containing the physical log file and/or the logical log files). >Other dbspaces (containing 'only' user data) need not be restored at >time of the level 0 cold restore. They can be restored later by doing a >warm restore (that however requires that you have backed up logical >logs and can restore them). >Saying "you have to restore the root dbspace ..." means that all chunks >that belong to root dbspace and critical dbspaces must be restored. As >you are talking about "chunks" I assume, that this was a chunk that >belonged to the root dbspace - as such it must be restored with the >level 0 cold restore. >Regards, >Martin **************************************************************************** *** Forum Note: Use "Reply" to post a response in the discussion forum. ******************************************************************************* Forum Note: Use "Reply" to post a response in the discussion forum.

I have checked and ONDBSPACEDOWN 2 # Dbspace down option: 0 = CONTINUE, 1 = ABORT, 2 = WAIT so i dont know why am i getting the error Opening primary chunks...Bad Primary Chunk '/informix/db/datadbs2_04'. succeeded Opening mirror chunks...succeeded Validating chunks...succeeded Initialize Async Log Flusher...succeeded Forking btree cleaner...succeeded Initializing DBSPACETEMP list...succeeded Checking database partition index...FAILED oninit: Fatal error in shared memory initialization Thankx vicky >Hi, >yes, Mark, that's what I was trying to say. >However, IDS can start and be on-line while a user >data dbspace is missing or corrupt. There is an >onconfig file parameter named ONDBSPACEDOWN. >IDS will continue when this is set to 0 (zero), it will >abort when this is set to 1 and it will wait (hang) when >this is set to 2. >So maybe on the system in question this is set to 1 >and that's why IDS aborts? >TIA, >Martin

Finally i reinitialized the test instance and started the restore using ontape -r. But the restore which was suppose to finish in 7-8 hrs on this server (as tested many times before)is not running over 15 hrs and i can see in onstat -D that one complete chunk is yet to be restored. onstat -D and onstat -u shows me that the restoration is in progress but dontknow the exact reason for this slow progress. My server is solaris 10 with 4cpu and 16 Gb RAM and the backups file is 50 GB(Zipped) using backup filter. I have restored this large backup on the dev server which is similiar to test server many times and it completed in 7-8 hrs. Pls let me know what should i check in order to find what is going wrong. Should i veryfiy the backup file. Dont know if Archecker works for ontape backup verification. Can there be some issue with the backup restore filter used. Pls help ! Many thanks vicky

Bet you three chocolate dipped donuts that your 'test' system has RAID5 disks under the chunks you created while the 'dev' system is using RAID10. Art On Fri, Aug 8, 2008 at 4:38 AM, VICKY H <vickykh26@gmail.com> wrote: > Finally i reinitialized the test instance and started the restore using > ontape > -r.> > But the restore which was suppose to finish in 7-8 hrs on this server (as > tested many times before)is not running over 15 hrs and i can see in onstat > -D > that one complete chunk is yet to be restored. > onstat -D and onstat -u shows me that the restoration is in progress but> dont > know the exact reason for this slow progress. > > My server is solaris 10 with 4cpu and 16 Gb RAM and the backups file is 50 > GB(Zipped) using backup filter. > > I have restored this large backup on the dev server which is similiar to > test > server many times and it completed in 7-8 hrs. > > Pls let me know what should i check in order to find what is going wrong. > Should i veryfiy the backup file. Dont know if Archecker works for ontape > backup verification. > Can there be some issue with the backup restore filter used. > > Pls help ! > > Many thanks > vicky > > > > ******************************************************************************* > Forum Note: Use "Reply" to post a response in the discussion forum. > > -- Art S. Kagel Oninit (www.oninit.com) IIUG Board of Directors (art@iiug.org) Disclaimer: Please keep in mind that my own opinions are my own opinions and do not reflect on my employer, Oninit, the IIUG, nor any other organization with which I am associated either explicitly or implicitly. Neither do those opinions reflect those of other individuals affiliated with any entity with which I am affiliated nor those of the entities themselves.

Art Kagel wrote: > Bet you three chocolate dipped donuts that your 'test' system has RAID5 > disks under the chunks you created while the 'dev' system is using RAID10. > Mmmm ... Dunkin Donuts .... mmmmmm.... arghlghgh -- Cheers, Obnoxio the Clown http://obotheclown.blogspot.com

One could envision an alternative career in law enforcement for Art after Oninit ... -----Original Message----- From: ids-bounces@iiug.org [mailto:ids-bounces@iiug.org] On Behalf Of Obnoxio The Clown Sent: Friday, August 08, 2008 10:33 AM To: ids@iiug.org Subject: Re: Test Restoration failed [13036] Art Kagel wrote: > Bet you three chocolate dipped donuts that your 'test' system has RAID5 > disks under the chunks you created while the 'dev' system is using RAID10. > Mmmm ... Dunkin Donuts .... mmmmmm.... arghlghgh -- Cheers, Obnoxio the Clown http://obotheclown.blogspot.com ************************************************************************ ******* Forum Note: Use "Reply" to post a response in the discussion forum.

Yeah, wait until he starts whacking you with that BAARF stick! (www.baarf.com) -- Bob -------------- Original message -------------- From: "Plugge, Joe R." <JRPlugge@west.com> > One could envision an alternative career in law enforcement for Art > after Oninit ... > > -----Original Message----- > From: ids-bounces@iiug.org [mailto:ids-bounces@iiug.org] On Behalf Of > Obnoxio The Clown > Sent: Friday, August 08, 2008 10:33 AM > To: ids@iiug.org > Subject: Re: Test Restoration failed [13036] > > Art Kagel wrote: > > Bet you three chocolate dipped donuts that your 'test' system has > RAID5 > > disks under the chunks you created while the 'dev' system is using > RAID10. > > > > Mmmm ... Dunkin Donuts .... mmmmmm.... arghlghgh > > -- > Cheers, > Obnoxio the Clown > > http://obotheclown.blogspot.com > > ************************************************************************ > ******* > Forum Note: Use "Reply" to post a response in the discussion forum. > > > ******************************************************************************* > Forum Note: Use "Reply" to post a response in the discussion forum. >