TimeSeries
Clive reported that IBM Informix Dynamic Server 11.70.FC1IE (Innovator-C/community edition) could be reliably crashed while experimenting with the TimeSeries datablade: an assertion failure in mtex.c line 498 (\\"No Exception Handler\\", MT_EX_OS/mem), killing the master daemon and panicking the server. Replies asked whether it followed an upgrade (it was a brand-new instance, with 11.70 auto-registering TimeSeries/Spatial) and requested the AF stack, OS version and a minimal reproducible test case. The analysis moved off-list, so no cause or fix is recorded in the thread.
Auto-generated by DrWatson from the posts below — may be imperfect; read the full thread.
IBM Informix Dynamic Server Version 11.70.FC1IE Software Serial Number
AAA#B000000
So I'm having a play with this and TS
I can reliably cause the engine to AF
11:14:37 stack trace for pid 24588 written to
/opt/IBM/informix/tmp/af.29de3252
11:14:37 Assert Failed: No Exception Handler
11:14:37 IBM Informix Dynamic Server Version 11.70.FC1IE
11:14:37 Who: Session(9619, informix@aa.bbb.ccc.dddddd, 27437, 0x4b80ff80)
Thread(9718, sqlexec, 4b7d2370, 1)
File: mtex.c Line: 498
11:14:37 Results: Exception Caught. Type: MT_EX_OS, Context: mem
11:14:37 Action: Please notify IBM Informix Technical Support.
11:14:37 See Also: /opt/IBM/informix/tmp/af.29de3252, shmem.29de3252.0
11:14:40 Starting crash time check of:
11:14:40 1. memory block headers
11:14:40 2. stacks
11:14:40 Crash time checking found no problems
11:14:40 mtex.c, line 498, thread 9718, proc id 24588, No Exception
Handler.
11:14:40 The Master Daemon Died
11:14:40 PANIC: Attempting to bring system down
As this is a 'community supported edition' who wants to see the af? :-)
--
Clive
On Dec 17, 11:58 am, Clive Eisen <cl...@serendipita.com> wrote:
> IBM Informix Dynamic Server Version 11.70.FC1IE Software Serial Number
> AAA#B000000
>
> So I'm having a play with this and TS
>
> I can reliably cause the engine to AF
>
> 11:14:37 stack trace for pid 24588 written to
> /opt/IBM/informix/tmp/af.29de3252
> 11:14:37 Assert Failed: No Exception Handler
> 11:14:37 IBM Informix Dynamic Server Version 11.70.FC1IE
> 11:14:37 Who: Session(9619, infor...@aa.bbb.ccc.dddddd, 27437, 0x4b80ff80)
> Thread(9718, sqlexec, 4b7d2370, 1)
> File: mtex.c Line: 498
> 11:14:37 Results: Exception Caught. Type: MT_EX_OS, Context: mem
> 11:14:37 Action: Please notify IBM Informix Technical Support.
> 11:14:37 See Also: /opt/IBM/informix/tmp/af.29de3252, shmem.29de3252.0
> 11:14:40 Starting crash time check of:
> 11:14:40 1. memory block headers
> 11:14:40 2. stacks
> 11:14:40 Crash time checking found no problems
> 11:14:40 mtex.c, line 498, thread 9718, proc id 24588, No Exception
> Handler.
> 11:14:40 The Master Daemon Died
> 11:14:40 PANIC: Attempting to bring system down>
> As this is a 'community supported edition' who wants to see the af? :-)
>
> --
> Clive
V11.70 probably doesn't help things. Did you follow the upgrade steps
religiously?
If you created a new database, register the blade and load data,
update stats etc then run your test, does it still fail. If so then
it's probably a bug - assuming it worked before.
If you upgraded then hit the problem then maybe you can fix.
So, if upgrade then have you got a stack trace and a list of functions
(or perhaps a dead master deamon isn't helping with a missing stack!)
and a bit more info Maybe the SQL that causes the fail as well. Not a
TS expert, and usually it's a deref pointer. Otherwise perhaps you can
trap the SQL upon error and get a stack this way for the session.
Can't promise I'll be around to answer, hope this is some help. Is
posting stacks in newsgroups allowed as it causes searching headaches,
be careful. Probably someone in DSD could run a 2 second stack lookup
for you.
IBM's site has been known to allow searches for some known issues.
Why is upgrading versions and blades such a headache. :>
I'm wrong, learnt something new ) OK, 11.70 automatically registers timeseries and spatial. Yeah, now I can see why you get an AF, especially if it's an upgrade. The documentation isn't very helpful either - http://publib.boulder.ibm.com/infocenter/idshelp/v117/topic/com.ibm.tms.doc/tms.htm
This is a brand new instance - no upgrade involved -- Clive On 20/12/2010 11:04, PeterP wrote: > I'm wrong, learnt something new ) > > OK, 11.70 automatically registers timeseries and spatial. Yeah, now I > can see why you get an AF, especially if it's an upgrade. > The documentation isn't very helpful either - > http://publib.boulder.ibm.com/infocenter/idshelp/v117/topic/com.ibm.tms.doc/tms.htm > _______________________________________________ > Informix-list mailing list > Informix-list@iiug.org > http://www.iiug.org/mailman/listinfo/informix-list
On Dec 17, 3:58 am, Clive Eisen <cl...@serendipita.com> wrote:
> IBM Informix Dynamic Server Version 11.70.FC1IE Software Serial Number
> AAA#B000000
>
> So I'm having a play with this and TS
>
> I can reliably cause the engine to AF
>
> 11:14:37 stack trace for pid 24588 written to
> /opt/IBM/informix/tmp/af.29de3252
> 11:14:37 Assert Failed: No Exception Handler
> 11:14:37 IBM Informix Dynamic Server Version 11.70.FC1IE
> 11:14:37 Who: Session(9619, infor...@aa.bbb.ccc.dddddd, 27437, 0x4b80ff80)
> Thread(9718, sqlexec, 4b7d2370, 1)
> File: mtex.c Line: 498
> 11:14:37 Results: Exception Caught. Type: MT_EX_OS, Context: mem
> 11:14:37 Action: Please notify IBM Informix Technical Support.
> 11:14:37 See Also: /opt/IBM/informix/tmp/af.29de3252, shmem.29de3252.0
> 11:14:40 Starting crash time check of:
> 11:14:40 1. memory block headers
> 11:14:40 2. stacks
> 11:14:40 Crash time checking found no problems
> 11:14:40 mtex.c, line 498, thread 9718, proc id 24588, No Exception
> Handler.
> 11:14:40 The Master Daemon Died
> 11:14:40 PANIC: Attempting to bring system down>
> As this is a 'community supported edition' who wants to see the af? :-)
>
> --
> Clive
Yeah, send the stack from the af as well as Operating System Version
and the simplest possible test case to reproduce it.
On 21/12/2010 01:06, david@smooth1.co.uk wrote: > Clive > > Yeah, send the stack from the af as well as Operating System Version > and the simplest possible test case to reproduce it. Many thanks for the offer David - I had a prior offer offlist Will revert when we understand what's happening -- Clive