Assert Failed: pthdrpage:ptaltdescs2:bad alter sl

Hi folks I have this recurring problem while loading data. This leads to a complete system hanger. I can't even use oncheck as requested, nor can I use onmode -ky to shutdown the system. The only way I can restart the informix is by rebooting the database server. Any Ideas ??? thanks a lot 11:18:37 Assert Failed: pthdrpage:ptaltdescs2:bad alter slot 11:18:37 Informix Dynamic Server Version 9.40.FC1 11:18:37 Who: Session(1658, tdfadm@hpsophia, 15410, c00000008e3bcbb8) Thread(2547, sqlexec, c00000008e37c918, 1) File: rspartn.c Line: 5718 11:18:37 Results: Cannot use TBLSpace page for TBLSpace 7348794 11:18:37 Action: Run 'oncheck -pt 7348794' 11:18:37 stack trace for pid 13802 written to /scratch/informix/DUMP/af.ddbd27d 11:18:37 See Also: /scratch/informix/DUMP/af.ddbd27d, shmem.ddbd27d.0 11:18:55 pthdrpage:ptaltdescs2:bad alter slot

Well have you ?? Run the suggested oncheck!! Check the DUMP files for more information, but you seem to have disk corruption and may need to unload around the errors, drop the chunks/spaces and rebuild. Keith On 21/11/06, MIGUEL DE BARROS <miguel.de.barros@nxp.com> wrote: > > Hi folks > > I have this recurring problem while loading data. > This leads to a complete system hanger. I can't even use oncheck as requested, > nor can I use onmode -ky to shutdown the system. The only way I can restart > the informix is by rebooting the database server. Any Ideas ??? > > thanks a lot > > 11:18:37 Assert Failed: pthdrpage:ptaltdescs2:bad alter slot > 11:18:37 Informix Dynamic Server Version 9.40.FC1 > 11:18:37 Who: Session(1658, tdfadm@hpsophia, 15410, c00000008e3bcbb8) > > Thread(2547, sqlexec, c00000008e37c918, 1) > > File: rspartn.c Line: 5718 > 11:18:37 Results: Cannot use TBLSpace page for TBLSpace 7348794 > 11:18:37 Action: Run 'oncheck -pt 7348794' > 11:18:37 stack trace for pid 13802 written to > /scratch/informix/DUMP/af.ddbd27d > 11:18:37 See Also: /scratch/informix/DUMP/af.ddbd27d, shmem.ddbd27d.0 > 11:18:55 pthdrpage:ptaltdescs2:bad alter slot > > > ******************************************************************************* > Forum Note: Use "Reply" to post a response in the discussion forum. > >

Hi, Can you execute the oncheck after reboot ? What is the output ? It seems you have a corrupted page ! Only IBM support can help here. Marcus -----Original Message----- From: ids-bounces@iiug.org [mailto:ids-bounces@iiug.org] On Behalf Of MIGUEL DE BARROS Sent: Tuesday, November 21, 2006 11:55 AM To: ids@iiug.org Subject: Assert Failed: pthdrpage:ptaltdescs2:bad alter sl [7844] Hi folks I have this recurring problem while loading data. This leads to a complete system hanger. I can't even use oncheck as requested, nor can I use onmode -ky to shutdown the system. The only way I can restart the informix is by rebooting the database server. Any Ideas ??? thanks a lot 11:18:37 Assert Failed: pthdrpage:ptaltdescs2:bad alter slot 11:18:37 Informix Dynamic Server Version 9.40.FC1 11:18:37 Who: Session(1658, tdfadm@hpsophia, 15410, c00000008e3bcbb8) Thread(2547, sqlexec, c00000008e37c918, 1) File: rspartn.c Line: 5718 11:18:37 Results: Cannot use TBLSpace page for TBLSpace 7348794 11:18:37 Action: Run 'oncheck -pt 7348794' 11:18:37 stack trace for pid 13802 written to /scratch/informix/DUMP/af.ddbd27d 11:18:37 See Also: /scratch/informix/DUMP/af.ddbd27d, shmem.ddbd27d.0 11:18:55 pthdrpage:ptaltdescs2:bad alter slot **************************************************************************** *** Forum Note: Use "Reply" to post a response in the discussion forum.

This is the result of oncheck after reboot. I also did a consistency check on the tables that were beeing loaded when the assert failed. ( onstat -g ses <ses id> ). There were no complaints. What I found out is the since some time the server was complaining about the physicallog beeing too small. Could this have been the reason for a corrupted page ? I m also veryfing with our system administrator on possible disk failures. [informix@hpsophia]: oncheck -pt 7348794 TBLspace Report for Unknown:Unknown.70223a ISAM error: no record found. ISAM error: no record found. Error opening TBLspace 70223a.

Several points: 1- Platform and IDS version information is missing from this post. That is often helpful. 2- What does onstat -g ses have to do with consistency checking the tables? 3- Did you run oncheck -cDI and oncheck -cr? Those are the standard checks you need to do first. 4- Where did you get the partnum for: oncheck -pt 7348794? Is that one that the log of .af file was complaining about? 5- Phys log too small! That's usually a harmless warning and there have been versions that erroneously complain about this under certain conditions (see why version info would help?) Art S. Kagel ----- Original Message ----- From: Miguel De Barros <ids@iiug.org> At: 11/21 8:24:25 This is the result of oncheck after reboot. I also did a consistency check on the tables that were beeing loaded when the assert failed. ( onstat -g ses <ses id> ). There were no complaints. What I found out is the since some time the server was complaining about the physicallog beeing too small. Could this have been the reason for a corrupted page ? I m also veryfing with our system administrator on possible disk failures. [informix@hpsophia]: oncheck -pt 7348794 TBLspace Report for Unknown:Unknown.70223a ISAM error: no record found. ISAM error: no record found. Error opening TBLspace 70223a. ******************************************************************************* Forum Note: Use "Reply" to post a response in the discussion forum.

Hi Art I used onstat - g ses to find out which sql command was running when the assert failed. I wanted to know which table was concerned. Yes I used oncheck -cDI -cr . No complaints. I performed the oncheck as requested in the online log file. The result is in my previous post. Right now I m loading again data, but without loading the data that were beeing loaded when the assert occurred. I'll wait until this is done. If nothing happens I must assume that the assert fail depends on a certain sql command. Miguel

I should apologize, I had not noticed that this was a followup to a previous post. It's good to keep the threaded history when you followup to avoid confusion though. How are you loading the data? The error from the oncheck -pt and the 'slot alter' error in the initial has me worried. I would seriously want to export all my data and initialize and reload a clean instance. I suspect serious corruption to the on-disk data structures. IB you have checked for physical corruption and error reports from your drives/arrays so that's out (isn't RAID5 is it?). Art S. Kagel ----- Original Message ----- From: Miguel De Barros <ids@iiug.org> At: 11/21 9:07:31 Hi Art I used onstat - g ses to find out which sql command was running when the assert failed. I wanted to know which table was concerned. Yes I used oncheck -cDI -cr . No complaints. I performed the oncheck as requested in the online log file. The result is in my previous post. Right now I m loading again data, but without loading the data that were beeing loaded when the assert occurred. I'll wait until this is done. If nothing happens I must assume that the assert fail depends on a certain sql command. Miguel ******************************************************************************* Forum Note: Use "Reply" to post a response in the discussion forum.

HI Art as I suspected it is an SQL command that is causing this type of crash :-( Our server log files did not indicate any HW failure. Thank you any way for helping :-) cheers Miguel