JBASE

[Srinivas G]

Hi All,
In jBASE journalling, for a distributed file, there are 2 entries as shown below.... 
 

ID....... jBASE..... Type....... Record Key..... Full file path name................ Update.. Update..... Total Log..

          login                                                                      Time     Date        Record Size

12*3*590898*12163773 kk1233 WRITE 000121897290872 /kkgt24/bnk/bnk.run/FBNK.STMT002 12:35:07 18 JUL 2008 504269

07                                1-20080723

7*3*628233*121637730 kk1233 WRITE 000121897290872 /kkgt24/bnk/bnk.run/../../bnk.dist/ 12:35:07 18 JUL 2008 504279

7                                  1-20080723     dist.sp/SP.Q.3

 

As we can see above, for the file STMT.PRINTED, there are 2 entries 1 for the MAIN file and the other for the PART file.... with LOGSIZE of 504269 & 504279.... which is quite huge... 

 

Does jBASE journalling work in this way or is there any way where we can aviod updates to the MAIN file i.e. FBNK.STMT002???

 

Regards,

KK 

 

 

 
 

 

 

 

 

[uiterwyk]

You need to shut off journaling for either the stub file or the
partition files.

If my memory serves, jBase suggests that you disable journaling for
the partition files.

I have always felt it was more effective to shut journaling of for the
stub file, because that way one saves the processing effort on the
backup machine of processing the partitioning algorithm subroutine.

Of course, shutting of journaling is done using the jchmod command.

HTH

Robert Uiterwyk

> KKIIf

[Srinivas G]

Thanks Robert for the quick reply... but when I use the jchmod command on the stub file... i.e. jchmod -L FBNK.STMT002, there is NO output... so how do I actually see that journalling is dis-abled for this stub file??? Also I can see some data on the stub file???

 

So, if i dis-able the stub file from logging, will this have any affect when I do a recovery???

 

[uiterwyk]

Hi KK,

To see the status of journaling (and a whole bunch of other info) do a
jstat -v [filename]

among other things you will see a LOG=[NO/YES]

For a large file, this may take a while to respond.

And the answer is no, it will not have any effect on recovery, AS LONG
AS YOU DON'T CHANGE THE PARTITIONING ALGORITHM.

If you think you might do that some day, then do it the "jBase way".
I.E. Log the stub file and turn of logging on the partition files.

Robert

[Srinivas G]

Hi Robert,

 

The file that am talking about is FBNK.STMT.PRINTED... this file has been distributed into 4 part files i.e. SP.Q1, SP.Q2, SP.Q3, SP.Q4 & the STUB file is FBNK.STMT002.

 

So, do you mean that I need to dis-able logging for the PART files & en-able for the STUB file alone... ??? My question is in TJ Logs, I can see 2 entries for distributed files... 1 for the PART file & 1 for the STUB file, with a huge amount of data.... 

 

Is there any way where I can have only 1 entry for distributed files without having any adverse affect while doing recovery??? This is reduce the data size on the logsets....

 

And am not changing the PARTITIONING ALGORITHM..... Also, I have many distributed files like STMT.ENTRY, CATEG.ENTRY, RE.CONSOL.SPEC.ENTRY, FBNK.FUNDS.TRANSFER$HIS etc... with different algorithms...

 

Many thanks,

KK

[uiterwyk]

Hi KK

IMHO, whar you should do is disable the logging on any STUB file for
the distributed files and make SURE that the PART files have logging
enabled.

You will then see that you only have one entry in the TJ for a
transaction, to the PART file, and not to the STUB file.

Having the two entries as you have now means that your backup machine
is doing twice the work, as it is writing each distributed file
transaction twice.

HTH.

Robert

[uiterwyk]

KK,

I just noticed how general the subject is. One should always try to
use a more specific subject for others reading the group. I have
changed it

Robert

[Srinivas G]

Thanks Robert for your support & help... from now on I will put on more meaningful subjects...

 

KK

[karthick kumar j]

Hi All,

 

I need to call a jbase subroutine from Java program. how i can implement it in JBASE can any one help me on this?

 

Regards,

Karthick