jBASE Journalling

[Road Jogger]

Hi There,

 

I got the below message when journalling was on....

 

Record 'DC082201066007' written to file 'FBNK.DC.BATCH.CONTROL'

FATAL ERROR: From user cs8732 at Thu Aug  7 15:36:39 2008
Process ID 10248194 , Port 1354 , tty /dev/pts/78
    Error 9 while writing to log journal

After I got this message, my session got hanged and I dis-connected the same... I was trying to change the above record from jBASE prompt when this happened... My jBASE release 4.1.5.16.

Any Ideas

 

Thanks

Narayan

[Road Jogger]

Hi All,

 

Can someone out there throw some light on this issue????

 

Regards,

Narayan

 

[Jim Idle]

On Mon, 2008-08-11 at 14:24 +0200, Road Jogger wrote:

Hi All,

 

Can someone out there throw some light on this issue????

 

Regards,

Narayan

 

On 8/8/08, Road Jogger <roadj...@gmail.com> wrote:

Hi There,

 

I got the below message when journalling was on....

 

Record 'DC082201066007' written to file 'FBNK.DC.BATCH.CONTROL'

FATAL ERROR: From user cs8732 at Thu Aug  7 15:36:39 2008
Process ID 10248194 , Port 1354 , tty /dev/pts/78
    Error 9 while writing to log journal

Error 9. is Bad file file descriptor on UNIX. However, you don't tell us what operating system this is and so on. You get this error if somehow a piece of code closes a file or otherwise plays with the C file descriptor and then tries to reuse it. I suppose it might happen if the file system was taken offline too. The log files are not on a network connection are they [this is a really bad idea of course ;-)]?

After I got this message, my session got hanged and I dis-connected the same... I was trying to change the above record from jBASE prompt when this happened... My jBASE release 4.1.5.16.

Operating system? jdiag output? Size of log files? Number of log files? Did you check and make sure that the filesysyem containing the log files is not full? PLease supply as mcuh information as possibler when you pose a question :-).

Jim

[Road Jogger]

Hi Jim,

 

The OS we use is AIX 5.3 (64 bit).... I did check the log file size and it was not full.... at that time when I got this error message,  the current log set was 3 & if I remember correctly, the size of the log set was around 19.2 GB.... i.e. each logset was around 1.2 GB... the logsets are NOT mounted on a network connection.... they reside on a separate volume group namely /t24tj/tj_log & inside this I have 2 directories namely

 

logsets

status

 

Under logsets, there are 4 sub-directories namely set1, set2, set3, set4....so each sub-directory (i.e. set1... etc...) contains the files for each logset....

 

Also, I suppose there is a command in jBASE where the error codes can be mapped... i.e. something like jmakeerr can you also let me know how to use this as well???

 

Do you need any other info????

 

Many Thanks

Narayan

[Jim Idle]

On Tue, 2008-08-12 at 09:32 +0200, Road Jogger wrote:

Hi Jim,

 

The OS we use is AIX 5.3 (64 bit).... I did check the log file size and it was not full.... at that time when I got this error message,  the current log set was 3 & if I remember correctly, the size of the log set was around 19.2 GB.... i.e. each logset was around 1.2 GB... the logsets are NOT mounted on a network connection.... they reside on a separate volume group namely /t24tj/tj_log & inside this I have 2 directories namely

 

logsets

status

 

Under logsets, there are 4 sub-directories namely set1, set2, set3, set4....so each sub-directory (i.e. set1... etc...) contains the files for each logset....

 

Also, I suppose there is a command in jBASE where the error codes can be mapped... i.e. something like jmakeerr can you also let me know how to use this as well???

ERRMSG nnnn

 

Do you need any other info????

Not really. There isn't much I can tell you other than this seems to be a transient error, most likely caused by a jBASE bug that is rare enough that it is not usually seen. All you can do is monitor the system and see if it comes up again. This error can come up when permissions are not quite right. Perhaps somewhere in that mix you did something like su - another user and then could not open the logs (this is just speculation). I have also seen these kind of transient errors from SAN arrays (worth checking that you have the latest 5.3 patches and the latest patches for any separate hardware.

Jim

[Tec Murho]

Do u still have same problem
if so can u use jrf and check about ur file if it needs downsizing or resizing
best regards

[Road Jogger]

I didnt get this problem again... but I suspect that I might have done a "SU" to the environment (Pointed out by Jim).... as I was having some 3-4 sessions at that time... but am not too sure...

 

There is no need to resize the file.... because thats the 1st thing I checked.... even the file permissions are also OK.... 775...

 

Anyway, journalling is on & till now there are no issues...lets see how it goes...

 

Thanks for asking

 

Narayan