Hi All,

Hopefully this is easy for you guys to help me out, I havent seen this issue posted anywhere as yet, but if I missed it I apologise for duplicating.

Basically when I start informix with ./oninit -iy(v) it all seems to work ok, I add databases and we all get along just smashingly.

I can shut this server down with ./onmode -yuk all ok on this front also. So you would think I have a nice working installation.

My problem is when I attempt to start up again after the shutdown with ./oninit -v i get the following error trace in the verbose output....

[Start Verbose Output]
ssyd0057dap:informix% ./oninit -v
Checking group membership to determine server run modesucceeded
Reading configuration file '/usr/informix//etc/onconfig'...succeeded
Creating /INFORMIXTMP/.infxdirs ... succeeded
Creating infos file "/usr/informix//etc/.infos.online1" ... "/usr/informix//etc/.conf.online1" ... succeeded
Writing to infos file ... succeeded
Checking config parameters...succeeded
Allocating and attaching to shared memory...succeeded
Creating resident pool 7082 kbytes...succeeded
Creating buffer pool 60002 kbytes...succeeded
Initializing rhead structure...succeeded
Initializing ASF ...succeeded
Initializing Dictionary Cache and SPL Routine Cache...succeeded
Bringing up ADM VP...succeeded
Creating VP classes...succeeded
Onlining 0 additional cpu vps...succeeded
Onlining 2 IO vps...succeeded
Initialization of Encryption...succeeded
Forking main_loop thread...succeeded
Initializing DR structures...succeeded
Forking 1 'ipcstr' listener threads...succeeded
Forking 1 'tlitcp' listener threads...succeeded
Forking 1 'tlitcp' listener threads...succeeded
Starting tracing...succeeded
Initializing 127 flushers...succeeded
Initializing log/checkpoint information...FAILED
oninit: Fatal error in shared memory initialization
[End Verbose Output]

My /etc/system looks like...

set shmsys:shminfo_shmmax=268435456
set semsys:seminfo_semmap=64
set semsys:seminfo_semmni=4096
set semsys:seminfo_semmns=4096
set semsys:seminfo_semmnu=4096
set semsys:seminfo_semume=64
set semsys:seminfo_semmsl=100
set shmsys:shminfo_shmmni=100
set shmsys:shminfo_shmseg=100
set shmsys:shminfo_shmmin=100

and from what I can see my onconfig and sqlhosts file are tiptop.

Any ideas at all from you guys would be appreciated.

Thanks in advance

Matt

Part of ‘oninit –iy’ process Informix will build sysmaster and sysutils database. Check the online log those databases was created successfully. Also, check for any other error messages in the online.log.

__
SC

Hi,

The sysmaster databases are created when I use the -i switch, and they are created successfully. I can create databases no problem with when its started this way.

included is an extract of the messages from the online.log

If you can think of anything else I would appreciate it thanks.


[LOG EXTRACT]
Tue Feb 8 13:06:24 2005

13:06:24 Event alarms enabled. ALARMPROG = '/usr/informix/etc/no_log.sh'
13:06:24 Booting Language <c> from module <>
13:06:24 Loading Module <CNULL>
13:06:24 Booting Language <builtin> from module <>
13:06:24 Loading Module <BUILTINNULL>
13:06:29 Dynamically allocated new virtual shared memory segment (size 8192KB)
13:06:29 IBM Informix Dynamic Server Version 9.40.UC3 Software Serial Number
13:06:30 Dynamically allocated new virtual shared memory segment (size 8192KB)
13:07:41 IBM Informix Dynamic Server Initialized -- Complete Disk Initialized.
13:07:41 WARNING: mt_aio_wait: errno == EINVAL
13:07:41 WARNING: mt_aio_wait: errno == EINVAL
13:07:41 I/O write chunk 1, pagenum 0, pagecnt 1 --> errno = 22
13:07:41 Checkpoint Completed: duration was 0 seconds.
13:07:41 Checkpoint loguniq 1, logpos 0xd0, timestamp: 226

13:07:41 Maximum server connections 0
13:07:41 Dataskip is now OFF for all dbspaces
13:07:41 On-Line Mode
13:07:41 Building 'sysmaster' database ...
13:07:51 Logical Log 1 Complete, timestamp:29544.
13:07:53 Booting Language <spl> from module <>
13:07:53 Loading Module <SPLNULL>
13:07:57 Unloading Module <SPLNULL>
13:07:57 Loading Module <SPLNULL>
13:08:02 Unloading Module <SPLNULL>
13:08:06 Loading Module <SPLNULL>
13:08:06 Logical Log 2 Complete, timestamp:48874.
13:08:12 Unloading Module <SPLNULL>
13:08:14 'sysmaster' database built successfully.
13:08:14 'sysutils' database built successfully.
13:08:14 'sysuser' database built successfully.
13:09:43 Shutdown Mode
13:09:44 Quiescent Mode
13:09:50 Checkpoint Completed: duration was 1 seconds.
13:09:50 Checkpoint loguniq 3, logpos 0x1f2018, timestamp: 58495
13:09:50 Maximum server connections 1
13:09:51 IBM Informix Dynamic Server Stopped.

13:10:15 IBM Informix Dynamic Server Started.
13:10:15 Requested shared memory segment size rounded from 67584KB to 81920KB
13:10:15 Segment locked: addr=0xa000000, size=83886080

Tue Feb 8 13:10:16 2005

13:10:16 Event alarms enabled. ALARMPROG = '/usr/informix/etc/no_log.sh'
13:10:16 Booting Language <c> from module <>
13:10:16 Loading Module <CNULL>
13:10:16 Booting Language <builtin> from module <>
13:10:16 Loading Module <BUILTINNULL>
13:10:21 Dynamically allocated new virtual shared memory segment (size 8192KB)
13:10:21 IBM Informix Dynamic Server Version 9.40.UC3 Software Serial Number
13:10:21 Dynamically allocated new virtual shared memory segment (size 8192KB)
13:10:22 I/O bad request chunk 1, pagenum 255263, pagecnt 1
13:10:22 IBM Informix Dynamic Server Stopped.

13:10:22 mt_shm_remove: WARNING: may not have removed all/correct segments
13:10:26 IBM Informix Dynamic Server Started.
13:10:26 Requested shared memory segment size rounded from 67584KB to 81920KB
13:10:27 Segment locked: addr=0xa000000, size=83886080
[/LOG EXTRACT]

Looks like you are getting error “I/O bad request chunk 1, pagenum 255263, pagecnt 1” every time informix trying to allocating 2nd virtual segments. I think it is something to do with HP-UX and number of shared memory segments. Check the informix release note if it mentioned any OS patch you need. As a workaround, increase the SHMVIRTSIZE in onconfig. The goal is database server should not allocate the 2nd virtual shared memory segment.

__
SC

Hi,

I am still having an issue with this. The troubling bit of this is the
"15:57:57 Logical Log File not found." message in the log.

What does this mean. Is this something to do with the raw device perhaps.

Any ideas?

Thanks

[LOG EXTRACT]
Wed Feb 16 15:57:51 2005

15:57:51 Event alarms enabled. ALARMPROG = '/usr/informix/etc/no_log.sh'
15:57:51 Booting Language <c> from module <>
15:57:51 Loading Module <CNULL>
15:57:51 Booting Language <builtin> from module <>
15:57:51 Loading Module <BUILTINNULL>
15:57:57 IBM Informix Dynamic Server Version 9.40.UC3 Software 15:57:57 Logical Log File not found.
15:57:57 IBM Informix Dynamic Server Stopped.

15:57:58 mt_shm_remove: WARNING: may not have removed all/correct segments
[/LOG EXTRACT]

Some more information. I was browsing through this forum and found someone having a similar problem. He was asked to post the af.* file as it was a dump of the server at time of crash.

I found that I have one of these and hopefully it makes things a little clearer for everyone that tries to help

16:11:00 mt_aio_wait(f68a5a0)
16:11:00 chunkio(3, 0xa724800, 0x0, 0, 0x1, 252762408, 34) errno 1
AIOREQ *a
0f86dcc4: 00000000 00000000 00000000 00000000 ........ ........
0f86dcd4: 00000000 00000000 0f68a778 0f68a3c8 ........ .h.x.h..
0f86dce4: 00000000 00000000 00000f75 41494f72 ........ ...uAIOr
0f86dcf4: 65710000 00000000 07d10000 00000000 eq...... ........
0f86dd04: 00000003 00000501 0000000b 00000002 ........ ........
0f86dd14: 00000001 00000000 00000001 0f68a620 ........ .....h.
0f86dd24: 00000000 00000000 00000016 00000001 ........ ........
0f86dd34: 00000000 00000000 00000000 ........ ....


Any ideas appreciated

Thanks

Did you check release notes? I remember a problem we had when we first upgraded to 9.40 The UNIX guy fixed it with the below changes per COMPAQ check your system release notes

3. Kernel Asynchronous I/O (KAIO)

On a heavily loaded machine, you may run out of KAIO resources. Refer
to the Compaq documentation to reconfigure the AIO resources used
by the kernel.

If you run out of KAIO resources, check the following kernel
configuration attributes.

aio_max_percent
aio_max_retry
aio_retry_scale
aio_task_max_num
sigqueue_max_num

Thanks for your replies guys, they ahev been very helpful. Although they have not seemed to solve the issue.

From time to time I am getting this error in the online log

11:49:55 Maximum server connections 16
11:55:21 WARNING: mt_aio_wait: errno == EINVAL
11:55:21 WARNING: mt_aio_wait: errno == EINVAL
11:55:21 I/O write chunk 1, pagenum 0, pagecnt 1 --> errno = 22
11:55:22 Fuzzy Checkpoint Completed: duration was 0 seconds, 88 buffers not flushed,
timestamp: 503325.
11:55:22 Checkpoint loguniq 63, logpos 0x65460, timestamp: 503325

The server has been started with ./oninit -iv and is running fine, and I am using the databases just fine. Again, I just worry about stopping the server as it wont restart without initialising the disk again which seems so very strange.

Any further help would be appreciated.

Can you include a copy of your onconfig file and the output from "onstat -d" if possible?