Re: Error in Enterprise Replication

Hi All, I have increased the number of logical logs by adding 9 more logical logs of 20MB each. Now I have 20 logical logs. But the problem is that after performing some table load activity all logs have become 100% used. At this stage, no application is running even I have restarted the IDS but previous logs are still showing 100% utilization of logical logs and no more space is available showing error "LONG TRANSACTION DETECTED". I will really appreciate if someone will help me in freeing this logical logs space. Test >onstat -l IBM Informix Dynamic Server Version 10.00.UC5 -- Quiescent -- Up 00:55:25 -- 46240 Kbytes Physical Logging Buffer bufused bufsize numpages numwrits pages/io P-1 6 8 8337 1196 6.97 phybegin physize phypos phyused %used 8:6931 5000 4627 6 0.12 Logical Logging Buffer bufused bufsize numrecs numpages numwrits recs/pages pages/io L-1 0 8 1326284 29680 4626 44.7 6.4 Subsystem numrecs Log Space used OLDRSAM 1326282 116667676 SBLOB 1 44 CDR 1 40 address number flags uniqid begin size used %used 502344b0 1 U-B---- 534 1:763 500 500 100.00 502344f8 2 U-B---- 535 1:1263 500 500 100.00 50234540 3 U-B---- 536 1:1763 500 500 100.00 50234588 4 U-B---- 537 1:2263 500 500 100.00 502345d0 7 U-B---- 538 8:923 500 500 100.00 50234618 5 U-B---- 539 1:2763 500 500 100.00 50234660 6 U-B---- 540 1:3263 500 500 100.00 502346a8 8 U-B---- 541 4:71993 500 500 100.00 502346f0 11 U-B---- 542 4:72993 500 500 100.00 50234738 9 U-B---- 543 4:72493 500 500 100.00 50234780 10 U-B---- 544 8:1431 500 500 100.00 502347c8 12 U-B---- 545 8:1931 5000 5000 100.00 50234810 13 U-B---- 546 8:11931 5000 5000 100.00 50234858 14 U-B---- 547 8:16931 5000 5000 100.00 502348a0 15 U---C-L 548 8:21931 5000 317 6.34 502348e8 16 U-B---- 530 8:26931 5000 5000 100.00 50234930 17 U-B---- 531 6:3 5000 5000 100.00 50234978 18 U-B---- 532 6:5003 5000 5000 100.00 502349c0 19 U-B---- 533 6:10003 5000 5000 100.00 604bbfb0 20 A------ 0 6:15003 5000 0 0.00 20 active, 20 total Test > Regards, Navaid Arif ----- Original Message ----- From: Navaid Arif To: admin-tools@iiug.org ; informix-list@iiug.org ; peoplemix@iiug.org ; classics@iiug.org ; ids@iiug.org Sent: Wednesday, May 02, 2007 11:36 AM Subject: Error in Enterprise Replication Hi All, I am using IDS 10 UC5 and during testing of Enterprise Replication facing following problem: Anyone have any idea about it? SQL: New Run Modify Use-editor Output Choose Save Info Drop Exit Modify the current SQL statements using the SQL editor. ----------------------- dbname@srvtest -------- Press CTRL-W for Help -------- load from loghisttbl.dmp insert into loghisttbl 458: Long transaction aborted. 12204: RSAM error: Long transaction detected. 847: Error in load file line 5825. Regards, Navaid Arif ------------------------------------------------------------------------------ _______________________________________________ Informix-list mailing list Informix-list@iiug.org http://www.iiug.org/mailman/listinfo/informix-list

On Wed, 16 May 2007 18:12:16 +0500, "Navaid Arif" <navaid.arif@access.net.pk> wrote: >Hi All, > I have increased the number of logical logs by adding 9 more logical logs of 20MB each. Now I have 20 logical logs. But the problem is that after performing some table load activity all logs have become 100% used. At this stage, no application is running even I have restarted the IDS but previous logs are still showing 100% utilization of logical logs and no more space is available showing error "LONG TRANSACTION DETECTED". > >I will really appreciate if someone will help me in freeing this logical logs space. > > Test >onstat -l > > IBM Informix Dynamic Server Version 10.00.UC5 -- Quiescent -- Up 00:55:25 -- 46240 Kbytes > > Physical Logging> Buffer bufused bufsize numpages numwrits pages/io > P-1 6 8 8337 1196 6.97 > phybegin physize phypos phyused %used > 8:6931 5000 4627 6 0.12 > > Logical Logging > Buffer bufused bufsize numrecs numpages numwrits recs/pages pages/io > L-1 0 8 1326284 29680 4626 44.7 6.4 > Subsystem numrecs Log Space used > OLDRSAM 1326282 116667676 > SBLOB 1 44 > CDR 1 40 > > address number flags uniqid begin size used %used > 502344b0 1 U-B---- 534 1:763 500 500 100.00 > 502344f8 2 U-B---- 535 1:1263 500 500 100.00 > 50234540 3 U-B---- 536 1:1763 500 500 100.00 > 50234588 4 U-B---- 537 1:2263 500 500 100.00 > 502345d0 7 U-B---- 538 8:923 500 500 100.00 > 50234618 5 U-B---- 539 1:2763 500 500 100.00 > 50234660 6 U-B---- 540 1:3263 500 500 100.00 > 502346a8 8 U-B---- 541 4:71993 500 500 100.00 > 502346f0 11 U-B---- 542 4:72993 500 500 100.00 > 50234738 9 U-B---- 543 4:72493 500 500 100.00 > 50234780 10 U-B---- 544 8:1431 500 500 100.00 > 502347c8 12 U-B---- 545 8:1931 5000 5000 100.00 > 50234810 13 U-B---- 546 8:11931 5000 5000 100.00 > 50234858 14 U-B---- 547 8:16931 5000 5000 100.00 > 502348a0 15 U---C-L 548 8:21931 5000 317 6.34 > 502348e8 16 U-B---- 530 8:26931 5000 5000 100.00 > 50234930 17 U-B---- 531 6:3 5000 5000 100.00 > 50234978 18 U-B---- 532 6:5003 5000 5000 100.00 > 502349c0 19 U-B---- 533 6:10003 5000 5000 100.00 > 604bbfb0 20 A------ 0 6:15003 5000 0 0.00 > 20 active, 20 total > Logical log usage doesn't seem to be a problem. The problem is that you are apparently trying to load too much data in a single transaction. Like David said, you can use dbload to set a commit point after every 'x' rows, which will make your life much easier. The problem with adding logs only to take care of a long transaction is that the exact number / size of logs to add isn't known. It's usually much easier to either use dbload or UNIX 'split' (assuming this is a UNIX box) to break up the data into smaller chunks to load. JWC