Hi,
activating AUTOSAVE crashes either DB Gui or complete database. I didn't look into this issue for 4 or 5 years because it is the same since I started with maxDB.
I tried this on the above mentioned software levels:
Database Studio is showing "<local>: <sid> Activating automatic log backup. --> for about 10 minutes. I can't stop this process, cancel or something else.
D:\sapdb\DBP\data\wrk\DBP\dbm.prt shows:
| 2013-07-16 09:37:05 | 3816 INF | 283 DBMSrv Command 'autolog_on ...' is being executed. |
| 2013-07-16 09:37:05 | 3816 INF | 1 DBMKnl Sending an administrative statement to the database DBP on computer localhost |
| 3816 INF | 8 DBMKnl Statement: AUTOSAVE ON '\\.\pipe\pipe_log'PIPE MEDIANAME 'ADSM_LOG' BLOCKSIZE 8 |
no entries in dbm.ebf or any other file.
Cancelling the request doesnt work in DB Studio. Closing DB Studio doesnt stop the processes, so I have to kill dbstudio.exe (javaw.exe).
Only killing the processes releases the client connection and allows me restarting DB Studio
| 2013-07-16 09:54:23 | 172 INF | 1 DBMSrvCo A DBM Server client connection was established at 2013-07-15 08:15:42 (client process has process ID 3076 on computer xxx). |
| 2013-07-16 09:54:23 | 172 ERR | -24811 DBMSrv ERR_CONNLOST: connection to client lost (error code 10) |
| 2013-07-16 09:54:23 | 172 INF | 2 DBMSrvCo A DBM Server client connection was released (client process has process ID 3076 on computer xxxx). |
Now I have the state "automatic log backup". But as far as I remember - the next time auto log backup tries to create a backup - the DB will shut down.
To reproduce the error I tried to restart automatic log backup with "every 10 minutes".
Now I get the error:
SQL error
-903,Host file I/O error
132,Cannot create medium on \\.\pipe\pipe_log with type Pipe
109,Medium on \\.\pipe\pipe_log already created as medium number 10
path : \\.\pipe\pipe_log
number : $MEDIA_NO$
is in use : 1
is opened : 1
ref count : 1
file handle : 0x9f0
task id : 97
even creating a manual backup doesn't work anymore "backup operation was unsuccessfulThe database was unable to fulfill a request (-903, Host file I/O error)."
After restarting the DB Logbackup is working again.
So if you have a good idea how to repair the combination tsm-adint-maxdb - I'd be very happy. But as this status didn't change for years I blamed IBM and looked for other solutions (schedule backups every xxx minutes via Windows-Scheduler + DB-Monitoring&Alerting).
Nonetheless a working event handler like we have in ASE would be nice.
Regards,
Peter