From nick.cook at itvision.com.au Tue Feb 28 06:06:09 2006 From: nick.cook at itvision.com.au (Nick Cook) Date: Tue Feb 28 07:05:09 2006 Subject: Jbase, locking and other issues. Message-ID: <1441FBD6BADE1C4289878CE45B036168D386C3@HATHOR.perth.itvision.com.au> We are in the process of duplicating our Universe database onto jBase with the specific intention of storing the data inside SQL using jEDI, and are going through quite an education process at the moment. Amongst various issues, one particular issue has come to light relating to locking, which differs markedly to Universe, Pick (and SQL come to that) Our application is very large, with about 750k lines of VB6 code, along with some 4,500 BASIC programs and subroutines. In short, a huge amount of legacy code and a considerable amount of newer code. In Universe, Pick etc if I lock a record no one else can write to that record until I release it. In jbase, if I lock a record with a READU statement, a write from any other user will only be stopped if it is preceded with a READU statement. My understanding of a lock on a record is that NO other user should be able to update that record until I release that lock, either by the original user writing to it, releasing it or using administrator commands to release locks. So if one user does a READU and locks a record, unless another user specifically attempts to lock the record then they may write to that record at will. I'm interested to know what other jBase users have done to stop this circumvention of normal locking rules. I appreciate that good practise dictates a readu before a write, however given the large amount of code that I have to deal with, I really dont want to go back and check that correct procedure is always followed on every write statement. Secondly, I'm having a lot of issues when creating the files in MSSQL via CREATE-FILE MYFILE TYPE=OLESQL. Whilst I've got a number of files across OK, (mostly simple files), files with a more complex structure (related Multi-values or Sub-values) tend to have issues, either immediately at the creation stage or with retrieving data once loaded. I'd be very interested to hear from anyone who has succeeded in porting a database to MSSQL using jBase and has any useful tips, tricks and advice they can throw my way. Incidentally I must credit my jBase supplier for the great support he provides, I'd be unhappy if he read this and felt it was a slight on the work he has done for us :-/ Regards, Nick Cook IT Vision -------------- next part -------------- An HTML attachment was scrubbed... URL: http://listserver.jbase.com/pipermail/tech/attachments/20060228/86bb3dee/attachment.htm From jose.a.desagun at credit-suisse.com Tue May 30 05:32:57 2006 From: jose.a.desagun at credit-suisse.com (De Sagun Jose A. (KSFE 41)) Date: Tue May 30 05:33:43 2006 Subject: jBASE: Unable to re-allocate 177408 bytes, errno = 12 at jediTransaction.c, 977(FATAL.ERROR, 440) Message-ID: <2F7E23A3629E2144B9924F6BF2B5836604AE9E4F@chsa1024.share.beluni.net> Hi There, Would like to ask if somebody already encountered the error message below. Please kindly advise. Details: _FIN.CYCLE.DATA_AL.POSITION.MERGE_3_24 MAY 2006_10:27:01_Single Thread routine AL.POSITION.MERGE _FIN.CYCLE.DATA_AL.POSITION.MERGE_3_24 MAY 2006_10:27:01_Updating the Locking with FIN.CYCLE.DATA-AL.POSITION.MERGE-1 and FBNK.JOB.LIST.1 _FIN.CYCLE.DATA_AL.POSITION.MERGE_3_24 MAY 2006_10:27:01_Session not doing SELECT processing 1 _FIN.CYCLE.DATA_AL.POSITION.MERGE_3_24 MAY 2006_10:27:01_Processing...1 Offset...0 Number of sessions...5 _FIN.CYCLE.DATA_AL.POSITION.MERGE_3_24 MAY 2006_11:11:33_SSELECT FBNK.POSITION Selected=640 time=0secs jBASE: Unable to re-allocate 177408 bytes, errno = 12 at jediTransaction.c,977(FATAL.ERROR,440) TRANSEND: Error 0 _FIN.CYCLE.DATA_AL.POSITION.MERGE_3_24 MAY 2006_11:11:39_11:11:39 Processed 1 lists [1] _FIN.CYCLE.DATA_AL.POSITION.MERGE_3_24 MAY 2006_11:11:39_SELECT FBNK.JOB.LIST.1 SAMPLE 100000 Selected=1 time=0secs _FIN.CYCLE.DATA_AL.POSITION.MERGE_3_24 MAY 2006_11:11:39_Processing...1 Offset...0 Number of sessions...5 _FIN.CYCLE.DATA_AL.POSITION.MERGE_3_24 MAY 2006_11:11:39_SSELECT FBNK.POSITION Selected=640 time=0secs jBASE: Unable to re-allocate 177920 bytes, errno = 12 at jediTransaction.c,977(EXCEPTION.LOG,69) TRANSEND: Error 0 _FIN.CYCLE.DATA_AL.POSITION.MERGE_3_24 MAY 2006_11:11:43_SELECT FBNK.JOB.LIST.1 SAMPLE 100000 Selected=1 time=0secs _FIN.CYCLE.DATA_AL.POSITION.MERGE_3_24 MAY 2006_11:11:43_Processing...1 Offset...0 Number of sessions...5 _FIN.CYCLE.DATA_AL.POSITION.MERGE_3_24 MAY 2006_11:11:43_SSELECT FBNK.POSITION Selected=640 time=0secs Process ID 696424 - Illegal instruction Program source name FATAL.ERROR , line 370 Thank you and kind regards, Jojo De Sagun CREDIT SUISSE IT International Application Management, KSFE 41 Binzm?hlestrasse 130 8050 Z?rich P.O. Box 600 CH-8070 Zurich Phone: +41.44.332.0520 (Zurich Office) +44.20.7.883.8170 (London Office) Fax: +41.44.332.0560 mailto:jose.a.desagun@credit-suisse.com www.credit-suisse.com -------------- next part -------------- An HTML attachment was scrubbed... URL: http://listserver.jbase.com/pipermail/tech/attachments/20060530/41d59792/attachment.htm From jose.a.desagun at credit-suisse.com Tue May 30 05:45:19 2006 From: jose.a.desagun at credit-suisse.com (De Sagun Jose A. (KSFE 41)) Date: Tue May 30 05:46:23 2006 Subject: Not enough space - ** Error [ SUBROUTINE_CALL_FAIL ] ** Message-ID: <2F7E23A3629E2144B9924F6BF2B5836604AE9E76@chsa1024.share.beluni.net> Hi There, May I ask if anybody encountered the error message below? Please kindly advise. Details: SC.BATCH.REP-SC.EXECUTED.ORDER-20 _SC.BATCH.REP_SC.EXECUTED.ORDER_5_12 MAY 2006_16:29:53_Updating the Locking with SC.BATCH.REP-SC.EXECUTED.ORDER-20 and FBNK.JOB.LIST.1 _SC.BATCH.REP_SC.EXECUTED.ORDER_5_12 MAY 2006_16:29:53_Starting job _SC.BATCH.REP_SC.EXECUTED.ORDER_5_12 MAY 2006_16:29:53_Processing...1 Offset...0 Number of sessions...5 jBASE: /jbglobus/grscsuk/t24/csut24/bnk/bnk.run/globuslib/lib376.so.15: Not enough space ** Error [ SUBROUTINE_CALL_FAIL ] ** Unable to perform CALL to subroutine SC.EXECUTED.ORDER , Line 595 , Source BAT CH.JOB.CONTROL Press C to continue or Q to quit Trap from an error message, error message name = SUBROUTINE_CALL_FAIL Line 595 , Source BATCH.JOB.CONTROL jBASE debugger->Q CONSOLIDATION.MARK NE 'C' Selected=1 time=0secs _SC.BATCH.REP_SC.EOD.EXCEPTION.LOG_5_12 MAY 2006_16:35:46_SSELECT F.BATCH BY BAT CH.STAGE Selected=279 time=0secs _SC.BATCH.REP_SC.POSITION.STMT_5_12 MAY 2006_16:35:46_Standard multi-thread job _SC.BATCH.REP_SC.POSITION.STMT_5_12 MAY 2006_16:35:46_Calling load routine jBASE: /jbglobus/grscsuk/t24/csut24/bnk/bnk.run/globuslib/lib387.so.16: Not enough space ** Error [ SUBROUTINE_CALL_FAIL ] ** Unable to perform CALL to subroutine SC.POSITION.STMT.LOAD , Line 768 , Source BATCH.JOB.CONTROL Press C to continue or Q to quit Trap from an error message, error message name = SUBROUTINE_CALL_FAIL Line 768 , Source BATCH.JOB.CONTROL jBASE debugger-> Thank you and kind regards, Jojo De Sagun CREDIT SUISSE IT International Application Management, KSFE 41 Binzm?hlestrasse 130 8050 Z?rich P.O. Box 600 CH-8070 Zurich Phone: +41.44.332.0520 (Zurich Office) +44.20.7.883.8170 (London Office) Fax: +41.44.332.0560 mailto:jose.a.desagun@credit-suisse.com www.credit-suisse.com -------------- next part -------------- An HTML attachment was scrubbed... URL: http://listserver.jbase.com/pipermail/tech/attachments/20060530/dc4bc0ce/attachment.htm