ER problem 9.3.UC2 / linux

Madison, I'm addressing you because the ONLY article I found on google pertaining to this problem was last answered by you, however in the last message you were asking for more info and that was in jan 2002. My problem is this: 21:53:55 Building 'syscdr' database ... 21:53:55 'syscdr' database built successfully. 21:53:56 CDR queuer initialization complete 21:53:56 CDR NIF listening on asf://g_billsvr1 21:54:47 Fuzzy Checkpoint Completed: duration was 0 seconds, 151 buffersnot flushed. 21:54:47 Checkpoint loguniq 103, logpos 0x32f52c 21:54:47 Maximum server connections 1 21:55:49 DDR Log Snooping - Snooping started in log 103 21:55:49 WARNING: mt_aio_wait: errno == EINVAL 21:55:49 I/O read chunk 3, pagenum 83717, pagecnt 32 --> errno = 22 21:56:03 CDR Grouper Fanout interface with the Dynamic Server has failed. 21:56:03 CDR Grouper FanOut thread is aborting. 21:56:04 DDR Log Snooping - Shutdown No matter how many times I stop and delete the servers and recreate them and try to start the replicates I am getting this problem. The odd thing is that the read chunk error happens on either the 3rd or 5th chunk same pagenum every time. I have checked the disk extensively and there are no errors or bad blocks, we even upgraded the firmware thinking it was the possibly the controller, but the same thing happens. Both the 3rd and 5th chunks are logical log dbspaces. This is a test platform so pretty much everything is on one disk. The primary server is running 9.3.UC2 and the target servers are all running 9.2.UC2 all on redhat 7.3. The last post on the previous guys problem you asked for onstat -d and onstat -l so I will post those as well. Any help you can offer isappreciated. Thanx, ~Tisha Shirah onstat -d output ----------------- Informix Dynamic Server Version 9.30.UC2 -- On-Line -- Up 1 days 13:00:48 -- 403092 Kbytes Dbspaces address number flags fchunk nchunks flags owner name 20a097d0 1 0x1 1 1 N informix rootdbs 20a78ec0 2 0x1 2 1 N informix physdbs 2133ce78 3 0x1 3 1 N informix llogdbs01 2133d018 4 0x1 4 1 N informix llogdbs02 2133d160 5 0x1 5 1 N informix llogdbs03 2133d2a8 6 0x1 6 1 N informix llogdbs04 2133d3f0 7 0x2001 7 1 N T informix tempdbs01 2133d538 8 0x2001 8 1 N T informix tempdbs02 2133d680 9 0x2001 9 1 N T informix tempdbs03 2133d7c8 10 0x2001 10 1 N T informix tempdbs04 2133d910 11 0x2001 11 1 N T informix tempdbs05 2133da58 12 0x2001 12 1 N T informix tempdbs06 2133dba0 13 0x2001 13 1 N T informix tempdbs07 2133dce8 14 0x2001 14 1 N T informix tempdbs08 2133de30 15 0x2001 15 1 N T informix tempdbs09 2133e018 16 0x2001 16 1 N T informix tempdbs10 2133e160 17 0x2001 17 1 N T informix tempdbs11 2133e2a8 18 0x2001 18 1 N T informix tempdbs12 2133e3f0 19 0x2001 19 1 N T informix tempdbs13 2133e538 20 0x2001 20 1 N T informix tempdbs14 2133e680 21 0x2001 21 1 N T informix tempdbs15 2133e7c8 22 0x1 22 2 N informix datadbs01 2133e910 23 0x1 23 2 N informix datadbs02 2133ea58 24 0x1 24 3 N informix datadbs03 2133eba0 25 0x1 28 1 N informix rddbs01 2133ece8 26 0x1 29 1 N informix rddbs02 2133ee30 27 0x1 30 1 N informix rddbs03 2133f018 28 0x1 31 4 N informix in_ldr01 2133f160 29 0x1 32 4 N informix in_ldr02 2133f2a8 30 0x1 33 4 N informix in_ldr03 2133f3f0 31 0x1 34 4 N informix in_ldr04 2133f538 32 0x1 35 4 N informix in_cldr01 2133f680 33 0x1 36 4 N informix in_cldr02 2133f7c8 34 0x1 37 4 N informix in_cldr03 2133f910 35 0x1 38 4 N informix out_ldr01 2133fa58 36 0x1 39 4 N informix out_ldr02 2133fba0 37 0x1 40 4 N informix out_ldr03 2133fce8 38 0x1 41 4 N informix out_ldr04 2133fe30 39 0x1 42 4 N informix out_ldr05 21340018 40 0x1 43 4 N informix out_ldr06 21340160 41 0x1 44 4 N informix out_ldr07 213402a8 42 0x1 45 4 N informix out_ldr08 213403f0 43 0x1 46 4 N informix out_cldr01 21340538 44 0x1 47 4 N informix out_cldr02 21340680 45 0x1 48 4 N informix out_cldr03 213407c8 46 0x1 49 4 N informix out_cldr04 21340910 47 0x1 50 4 N informix out_cldr05 21340a58 48 0x1 51 4 N informix out_cldr06 21340ba0 49 0x1 52 16 N informix cdrdbs01 21340ce8 50 0x1 89 16 N informix cdrdbs03 21340e30 51 0x1 126 16 N informix cdrdbs02 21341018 52 0x1 163 16 N informix cdrdbs04 21341160 53 0x1 179 1 N informix cdrdbs00 213412a8 54 0x1 181 1 N informix sendqdbs 213413f0 55 0x8001 182 1 N S informix sendqdatadbs 21341538 56 0x1 183 1 N informix in_cldr_pk 21341680 57 0x1 184 1 N informix in_ldr_pk 213417c8 58 0x1 185 1 N informix out_cldr_pk 21341910 59 0x1 186 1 N informix out_ldr_pk 21341a58 60 0x1 187 1 N informix in_ldr_uq 21341ba0 61 0x1 188 1 N informix out_ldr_uq 61 active, 2047 maximum Chunks address chk/dbs offset size free bpages flags pathname 20a09918 1 1 0 196608 191634 PO- /dev/chunks/rootdbs 20a78d50 2 2 196608 196608 0 PO- /dev/chunks/physdbs 21326018 3 3 0 1045875 627502 PO- /dev/chunks/llogdbs01 21326188 4 4 0 1045875 627502 PO- /dev/chunks/llogdbs02 213262f8 5 5 0 1045875 627502 PO- /dev/chunks/llogdbs03 21326468 6 6 0 1045875 627502 PO- /dev/chunks/llogdbs04 213265d8 7 7 0 1045875 1045822 PO- /dev/chunks/tempdbs01 21326748 8 8 0 1045875 1045822 PO- /dev/chunks/tempdbs02@@

Please open a case on this. I think that we found out that the fundamental problem was the required memory alignment for raw IO on Linux. "Tisha Shirah" <lshirah@simplified-nospam.com> wrote in message news:DgfQa.867$on5.328@newssvr23.news.prodigy.com... > Madison, > > I'm addressing you because the ONLY article I found on google pertaining to > this problem was last answered by you, however in the last message you were > asking for more info and that was in jan 2002. My problem is this: > > 21:53:55 Building 'syscdr' database ... > 21:53:55 'syscdr' database built successfully. > 21:53:56 CDR queuer initialization complete > 21:53:56 CDR NIF listening on asf://g_billsvr1 > 21:54:47 Fuzzy Checkpoint Completed: duration was 0 seconds, 151 buffers> not flushed. > 21:54:47 Checkpoint loguniq 103, logpos 0x32f52c > > 21:54:47 Maximum server connections 1 > 21:55:49 DDR Log Snooping - Snooping started in log 103 > 21:55:49 WARNING: mt_aio_wait: errno == EINVAL > 21:55:49 I/O read chunk 3, pagenum 83717, pagecnt 32 --> errno = 22 > 21:56:03 CDR Grouper Fanout interface with the Dynamic Server has failed. > 21:56:03 CDR Grouper FanOut thread is aborting. > 21:56:04 DDR Log Snooping - Shutdown> > No matter how many times I stop and delete the servers and recreate them and > try to start the replicates I am getting this problem. The odd thing is > that the read chunk error happens on either the 3rd or 5th chunk same > pagenum every time. I have checked the disk extensively and there are no > errors or bad blocks, we even upgraded the firmware thinking it was the > possibly the controller, but the same thing happens. Both the 3rd and 5th > chunks are logical log dbspaces. This is a test platform so pretty much > everything is on one disk. The primary server is running 9.3.UC2 and the > target servers are all running 9.2.UC2 all on redhat 7.3. > > The last post on the previous guys problem you asked for onstat -d and > onstat -l so I will post those as well. Any help you can offer is> appreciated. > > Thanx, > > ~Tisha Shirah > > onstat -d output -----------------> > Informix Dynamic Server Version 9.30.UC2 -- On-Line -- Up 1 days > 13:00:48 -- 403092 Kbytes > > Dbspaces > address number flags fchunk nchunks flags owner name > 20a097d0 1 0x1 1 1 N informix rootdbs > 20a78ec0 2 0x1 2 1 N informix physdbs > 2133ce78 3 0x1 3 1 N informix llogdbs01 > 2133d018 4 0x1 4 1 N informix llogdbs02 > 2133d160 5 0x1 5 1 N informix llogdbs03 > 2133d2a8 6 0x1 6 1 N informix llogdbs04 > 2133d3f0 7 0x2001 7 1 N T informix tempdbs01 > 2133d538 8 0x2001 8 1 N T informix tempdbs02 > 2133d680 9 0x2001 9 1 N T informix tempdbs03 > 2133d7c8 10 0x2001 10 1 N T informix tempdbs04 > 2133d910 11 0x2001 11 1 N T informix tempdbs05 > 2133da58 12 0x2001 12 1 N T informix tempdbs06 > 2133dba0 13 0x2001 13 1 N T informix tempdbs07 > 2133dce8 14 0x2001 14 1 N T informix tempdbs08 > 2133de30 15 0x2001 15 1 N T informix tempdbs09 > 2133e018 16 0x2001 16 1 N T informix tempdbs10 > 2133e160 17 0x2001 17 1 N T informix tempdbs11 > 2133e2a8 18 0x2001 18 1 N T informix tempdbs12 > 2133e3f0 19 0x2001 19 1 N T informix tempdbs13 > 2133e538 20 0x2001 20 1 N T informix tempdbs14 > 2133e680 21 0x2001 21 1 N T informix tempdbs15 > 2133e7c8 22 0x1 22 2 N informix datadbs01 > 2133e910 23 0x1 23 2 N informix datadbs02 > 2133ea58 24 0x1 24 3 N informix datadbs03 > 2133eba0 25 0x1 28 1 N informix rddbs01 > 2133ece8 26 0x1 29 1 N informix rddbs02 > 2133ee30 27 0x1 30 1 N informix rddbs03 > 2133f018 28 0x1 31 4 N informix in_ldr01 > 2133f160 29 0x1 32 4 N informix in_ldr02 > 2133f2a8 30 0x1 33 4 N informix in_ldr03 > 2133f3f0 31 0x1 34 4 N informix in_ldr04 > 2133f538 32 0x1 35 4 N informix in_cldr01 > 2133f680 33 0x1 36 4 N informix in_cldr02 > 2133f7c8 34 0x1 37 4 N informix in_cldr03 > 2133f910 35 0x1 38 4 N informix out_ldr01 > 2133fa58 36 0x1 39 4 N informix out_ldr02 > 2133fba0 37 0x1 40 4 N informix out_ldr03 > 2133fce8 38 0x1 41 4 N informix out_ldr04 > 2133fe30 39 0x1 42 4 N informix out_ldr05 > 21340018 40 0x1 43 4 N informix out_ldr06 > 21340160 41 0x1 44 4 N informix out_ldr07 > 213402a8 42 0x1 45 4 N informix out_ldr08 > 213403f0 43 0x1 46 4 N informix out_cldr01 > 21340538 44 0x1 47 4 N informix out_cldr02 > 21340680 45 0x1 48 4 N informix out_cldr03 > 213407c8 46 0x1 49 4 N informix out_cldr04 > 21340910 47 0x1 50 4 N informix out_cldr05 > 21340a58 48 0x1 51 4 N informix out_cldr06 > 21340ba0 49 0x1 52 16 N informix cdrdbs01 > 21340ce8 50 0x1 89 16 N informix cdrdbs03 > 21340e30 51 0x1 126 16 N informix cdrdbs02 > 21341018 52 0x1 163 16 N informix cdrdbs04 > 21341160 53 0x1 179 1 N informix cdrdbs00 > 213412a8 54 0x1 181 1 N informix sendqdbs > 213413f0 55 0x8001 182 1 N S informix > sendqdatadbs > 21341538 56 0x1 183 1 N informix in_cldr_pk > 21341680 57 0x1 184 1 N informix in_ldr_pk > 213417c8 58 0x1 185 1 N informix out_cldr_pk > 21341910 59 0x1 186 1 N informix out_ldr_pk > 21341a58 60 0x1 187 1 N informix in_ldr_uq > 21341ba0 61 0x1 188 1 N informix out_ldr_uq > 61 active, 2047 maximum > > Chunks > address chk/dbs offset size free bpages flags pathname > 20a09918 1 1 0 196608 191634 PO- > /dev/chunks/rootdbs > 20a78d50 2 2 196608 196608 0

My line manager has been doing some experiments with ER on Linux and this may be the problem he experienced. He logged a case with IBM Informix technical support on this issue. Their response is that there is an issue with raw IO and logical logs on Linux. Using cooked files for the logical logs works around the problem. From memory we have been told the bug will be fixed in 9.40.UC2. Ben.

Madison and Ben, Thanx for the help. I ended up removing the raw devices and creating cooked chunks for the logical logs and replication started fine. So is it a bug in the linux kernel or Informix? ~Tisha "Ben Thompson" <ben@nomonitorsoftspamthanks.com> wrote in message news:3f12a8f3$1@news.userve.net... > My line manager has been doing some experiments with ER on Linux and > this may be the problem he experienced. He logged a case with IBM > Informix technical support on this issue. Their response is that there > is an issue with raw IO and logical logs on Linux. Using cooked files > for the logical logs works around the problem. From memory we have been > told the bug will be fixed in 9.40.UC2. > > Ben. > >