Backup Level 1 blocking smart blob

Hi, We have set up a relatively large database containing smart blob data (> 500 GB). There is 3 tables in the DB, which is the only one in the instance, all with some "standard" columns and one blob column in a separate sblob dbspace, which is spread over 3 chunks. We are running IDS 11.70FC5GE on Linux (Ubuntu 10.04, 64bit) with attached external disk arrays (Fibre channel 4GBit, Raid 1). Chunks are in raw partitions from lvm2. We have tried to create level 1 backups using ontape. (weekly a level 0 and daily a level 1). The effect was that the DB was completly blocked for about an hour before the real level 1 backup started. There was not a trace in online.log that ontape was even started, but it was running and blocked the DB (maybe only the table, we have not done deep research on this issue). After the backup started (printed in online log that a backup was started), the DB was accessible again. Level 0 backup starts relatively immediately (after a checkpoint, which is common for all backups), and does not lock the DB longer than the checkpoint duration. Anybody encountered this before ? Any hints what to do ? We have opened a case, but the reseller did not give an reasonable hint up til now. We should execute onchecks for sblob, but these too blocked the DB, so we have to do that at maybe on a secondinstance started on a snapshot to make sure we do not lock database in production. Thank you for any thoughts on this. Marcus Haarmann

Open a PMR. This may be a known issue (I haven't see it, but...) and a fix may already be available in 11.70.FC6 or .FC7. Art Art S. Kagel Advanced DataTools (www.advancedatatools.com) Blog: http://informix-myview.blogspot.com/ Disclaimer: Please keep in mind that my own opinions are my own opinions and do not reflect on my employer, Advanced DataTools, the IIUG, nor any other organization with which I am associated either explicitly, implicitly, or by inference. Neither do those opinions reflect those of other individuals affiliated with any entity with which I am affiliated nor those of the entities themselves. On Fri, Feb 1, 2013 at 10:13 AM, Marcus Haarmann <marcus.haarmann@midoco.de>wrote: > Hi, > > We have set up a relatively large database containing smart blob data (> > 500 > GB). > There is 3 tables in the DB, which is the only one in the instance, all > with > some "standard" > columns and one blob column in a separate sblob dbspace, which is spread > over 3 chunks. > We are running IDS 11.70FC5GE on Linux (Ubuntu 10.04, 64bit) with attached > external disk arrays (Fibre channel 4GBit, Raid 1). > Chunks are in raw partitions from lvm2. > We have tried to create level 1 backups using ontape. (weekly a level 0 and > daily a level 1). > > The effect was that the DB was completly blocked for about an hour before > the > real level 1 backup started. > There was not a trace in online.log that ontape was even started, but it > was > running and blocked the DB > (maybe only the table, we have not done deep research on this issue). After > the backup started (printed in online log > that a backup was started), the DB was accessible again. > Level 0 backup starts relatively immediately (after a checkpoint, which is > common for all backups), > and does not lock the DB longer than the checkpoint duration. > > Anybody encountered this before ? Any hints what to do ? > We have opened a case, but the reseller did not give an reasonable hint up > til > now. > We should execute onchecks for sblob, but these too blocked the DB, > so we have to do that at maybe on a secondinstance started on a snapshot > to make sure we do not lock database in production. > > Thank you for any thoughts on this. > > Marcus Haarmann > > > > ******************************************************************************* > Forum Note: Use "Reply" to post a response in the discussion forum. > > --e0cb4efe33a82bae4f04d4abd93f

Hi Art, we already have 11.70FC7 here, so we will try to upgrade and then perform the operation again. But since the instance is used mostly 24x7, we have to announce a downtime first. I was just thinking anybody out there had this issue before and solved it ... Marcus ----- Ursprüngliche Mail ----- Von: "Art Kagel" <art.kagel@gmail.com> An: ids@iiug.org Gesendet: Freitag, 1. Februar 2013 17:01:00 Betreff: Re: Backup Level 1 blocking smart blob [29457] Open a PMR. This may be a known issue (I haven't see it, but...) and a fix may already be available in 11.70.FC6 or .FC7. Art Art S. Kagel Advanced DataTools (www.advancedatatools.com) Blog: http://informix-myview.blogspot.com/ Disclaimer: Please keep in mind that my own opinions are my own opinions and do not reflect on my employer, Advanced DataTools, the IIUG, nor any other organization with which I am associated either explicitly, implicitly, or by inference. Neither do those opinions reflect those of other individuals affiliated with any entity with which I am affiliated nor those of the entities themselves. On Fri, Feb 1, 2013 at 10:13 AM, Marcus Haarmann <marcus.haarmann@midoco.de>wrote: > Hi, > > We have set up a relatively large database containing smart blob data (> > 500 > GB). > There is 3 tables in the DB, which is the only one in the instance, all > with > some "standard" > columns and one blob column in a separate sblob dbspace, which is spread > over 3 chunks. > We are running IDS 11.70FC5GE on Linux (Ubuntu 10.04, 64bit) with attached > external disk arrays (Fibre channel 4GBit, Raid 1). > Chunks are in raw partitions from lvm2. > We have tried to create level 1 backups using ontape. (weekly a level 0 and > daily a level 1). > > The effect was that the DB was completly blocked for about an hour before > the > real level 1 backup started. > There was not a trace in online.log that ontape was even started, but it > was > running and blocked the DB > (maybe only the table, we have not done deep research on this issue). After > the backup started (printed in online log > that a backup was started), the DB was accessible again. > Level 0 backup starts relatively immediately (after a checkpoint, which is > common for all backups), > and does not lock the DB longer than the checkpoint duration. > > Anybody encountered this before ? Any hints what to do ? > We have opened a case, but the reseller did not give an reasonable hint up > til > now. > We should execute onchecks for sblob, but these too blocked the DB, > so we have to do that at maybe on a secondinstance started on a snapshot > to make sure we do not lock database in production. > > Thank you for any thoughts on this. > > Marcus Haarmann > > > > ******************************************************************************* > Forum Note: Use "Reply" to post a response in the discussion forum. > > --e0cb4efe33a82bae4f04d4abd93f ******************************************************************************* Forum Note: Use "Reply" to post a response in the discussion forum.

This is a problem on IDS. Occurs only on Level 1 Ontape and is related to SmartBlobs data. Informix Lab is already working on it, but still not a clear cause and, of course, not a solution. Maybe you can try to open a new PMR (another one is already created) trying to push Informix Team for a solution.