Number of logical logs growing ?

Hi all, I'm doing some tests on ( new for me ) old 12.10.FC9IE I have configured it on a small machine, and created a couple of small ( less than 1G ) databases. I've then moved logical log from rootdbs to a dedicated chunk, and so far so good. What I'm noticing is that the number of logical logs keeps growing also if the machine ( and the databases ) are doing nothing. I'll explain better: my instance is running on a machine to which I connect only when I'm back from work, and baically I do not operate on the DB. I've started the db with 12 logical logs, as I was used on my old 7.x DB, and after some time I found ( as now for example ) that there are 56 logical logs. No application is working on it, there are no application using the db and I have also modified the onconfig file to have the DINAMYC_LOGS set to 0 The only operation to DB is a crontab that I used to have also on my old 7.x DBs that, every 5 minutes issues an "ontape -a" In my previous experience with IDS this was needed to avoid long transation hanging and I have done the same here. The only real difference is that on old IDS the logs were in rootdbs while here I have dedicated a dbspace to them. The Logical log on this IDS instance are quite empty: ... ... ... 47fd3f38 29 U-B---- 44704 3:42053 1500 19 1.27 47fe7fa8 30 U-B---- 44705 3:43553 1500 9 0.60 48021fb0 31 U-B---L 44706 3:45053 1500 7 0.47 47faafb0 32 U-B---- 44707 3:46553 1500 5 0.33 47fd3fa0 33 U-B---- 44708 3:48053 1500 5 0.33 477374c8 34 U-B---- 44709 3:49553 1500 98 6.53 ... ... ... Is there any way to avoid this logs automatically grow ( in number ) ? I'm scared that they will, at a time, fill the dbspace and block some transaction. Thanks in advance Pierluigi

Are there any messages on the online log indicating when the logical logs are being added? What is the output of: onstat -g cfg DYNAMIC_LOGS onstat -g cfg LTXHWM onstat -g cfg LTXEHWM And just to be sure, what is the result of the following query on the sysmaster database: select * from sysconfig where cf_name = 'DYNAMIC_LOGS';

I can't find any interesting info on logs, but decided to go in another way. I've removed 40 logs, stopped the db, cleared the logfile and started again. I'm now tracking when it will add a log ( in a script ) and tail the latest lines of log file ( ol_informix1210.log ) t try to get a pattern. Here the output you requested: informix:~# onstat -g cfg DYNAMIC_LOGS IBM Informix Dynamic Server Version 12.10.FC9IE -- On-Line -- Up 3 days 02:01:17 -- 185992 Kbytes name current value DYNAMIC_LOGS 0 informix:~# onstat -g cfg LTXHWM IBM Informix Dynamic Server Version 12.10.FC9IE -- On-Line -- Up 3 days 02:01:29 -- 185992 Kbytes name current value LTXHWM 50 informix:~# onstat -g cfg LTXEHWM IBM Informix Dynamic Server Version 12.10.FC9IE -- On-Line -- Up 3 days 02:01:34 -- 185992 Kbytes name current value LTXEHWM 60 informix:~# echo "database sysmaster; select * from sysconfig where cf_name = 'DYNAMIC_LOGS' " | dbaccess Database selected. cf_id 23 cf_name DYNAMIC_LOGS cf_flags 163906 cf_original 0 cf_effective 0 cf_default 2 1 row(s) retrieved. Database closed. informix:~#

This is what I've found in logs at the time of crontab ontape -a jobs ( no one was connected to the db, neither no jobs were usign it ): 17:20:01 Logical Log 47416 - Backup Started 17:20:01 Logical Log 47416 - Backup Completed 17:20:01 Logical Log 47417 - Backup Started 17:20:01 Logical Log 47417 - Backup Completed 17:20:01 Logical Log 47418 - Backup Started 17:20:01 Logical Log 47418 - Backup Completed 17:20:01 Logical Log 47419 - Backup Started 17:20:01 Logical Log 47419 - Backup Completed 17:20:01 Logical Log 47420 - Backup Started 17:20:01 Logical Log 47420 - Backup Completed 17:20:01 Performance Advisory: The logical log is running out of room during checkpoint processing. 17:20:01 Results: Transactions are being blocked until the checkpoint is complete. 17:20:01 Action: Increase the logical log space size. 17:20:01 Logical Log 47421 - Backup Started 17:20:01 Checkpoint Completed: duration was 0 seconds. 17:20:01 Wed Jun 12 - loguniq 47421, logpos 0x18, timestamp: 0xf964e40 Interval: 2404 17:20:01 Maximum server connections 1 17:20:01 Checkpoint Statistics - Avg. Txn Block Time 0.000, # Txns blocked 1, Plog used 1316, Llog used 3840 17:20:01 Logical Log 47421 - Backup Completed 17:20:01 Logical Log 47422 - Backup Started 17:20:01 Logical Log 47422 - Backup Completed 17:20:01 Logical Log 47418 Complete, timestamp: 0xf964e86. 17:20:01 Logical Log 47419 Complete, timestamp: 0xf964e86. 17:20:01 Logical Log 47420 Complete, timestamp: 0xf964e86. 17:20:01 Logical Log 47421 Complete, timestamp: 0xf964e86. 17:20:01 Logical Log 47422 Complete, timestamp: 0xf964e86. 17:20:01 Log file 13 added to DBspace 3. 17:20:01 Checkpoint Completed: duration was 0 seconds. 17:20:01 Wed Jun 12 - loguniq 47423, logpos 0xa400, timestamp: 0xf964ee3 Interval: 2405 17:20:01 Maximum server connections 1 17:20:01 Checkpoint Statistics - Avg. Txn Block Time 0.001, # Txns blocked 0, Plog used 47, Llog used 22 17:20:01 ** AUTO TUNING - Logical Log extended. Any idea ?

Well, something is causing activity in the instance and consuming logical logs. The timestamps and ordering of the online log seem strange. All those entries on the same second, the backup start entries before the corresponding logical log completing. Is the machine hibernating? Or aome task in the Informix internal scheduler? Maybe turning on SQLTRACE and check if it is capturing any activity? Activating the audit log could also give some information. The addition of logical logs is probably because you must have the parameter AUTO_LLOG activated. According to the online documentation, it is activated if we choose to initialize an instance during the Informix installation.

Hi louis, I0ve managed to unconfigure the AUTO_LLOG, and now my growing problem is no more present. Thanks a lot for that info, I wouldn't be able to find it !