Upgrade from 10.00.FC6 - 11.70.FC1 ~ Answered: amber (solid confidence) — After a multi-day conversion-failure saga (dbspace free-space shortfall, then an RSAM component version-mismatch abort), the poster's own final post-mortem attributes the real root cause to undersized Unix kernel limits (nfile, semmni, etc.) for 11.70's higher process count and reports the specific values used, but never explicitly confirms full success.

Advisory only.

We are trying to upgrade our IDS engine in HP-UX B.11.11 from version 10.00.FC6 to version 11.70.FC1. We encountered error during the conversion. Please take a look at the error. Do we need to drop the sysmaster table from the old version before the conversion? Do we need a full backup from the old version for the conversion to run? 09:44:53 Conv/rev: Finished conversion for component SQL to component version 33 09:44:53 The dummy updates succeeded while converting database sysuser. 09:44:53 Internal Conversion Completed Successfuly 09:44:53 Conversion Enabling Client Connections 09:44:53 Conv/rev: Finished conversion for component SQL 09:44:53 Conv/rev: Started conversion for component SMI 09:44:53 *** SMI 1 11.70xC1 *** 09:44:53 Build of sysmaster and sysutils databases Started 09:44:53 Converting/Rebuilding 'sysmaster' database ... 09:44:54 Single-User Mode. 09:45:23 Process exited with return code 1: /bin/sh /bin/sh -c /apps/informix/etc/no_log.sh 2 23 "" "Logical Log 268859 Complete, timestamp: 0x14b87b4d." "" 23001 09:45:43 Process exited with return code 1: /bin/sh /bin/sh -c /apps/informix/etc/no_log.sh 2 23 "" "Logical Log 268860 Complete, timestamp: 0x14cc49bb." "" 23001 09:49:03 Checkpoint Completed: duration was 25 seconds. 09:49:03 Wed Jan 12 - loguniq 268861, logpos 0x3d31018, timestamp: 0x14d48991 Interval: 6 09:49:03 Maximum server connections 0 09:49:03 Checkpoint Statistics - Avg. Txn Block Time 0.000, # Txns blocked 0, Plog used 13312, Llog used 43437 09:54:11 Error building 'sysmaster' database. 09:54:11 See '/tmp/buildsmi.15000'. 09:54:11 Conv/rev: Finished conversion for component SMI to component version 2 09:54:11 Conv/rev: Finished conversion for component SMI 09:54:11 Conv/rev: Started conversion for component BAR 09:54:11 *** BAR 1 11.10xC1 *** 09:54:11 Conv/rev: Finished conversion for component BAR to component version 3 09:54:11 *** BAR 1 11.50xC1 *** 09:54:11 Conv/rev: Finished conversion for component BAR to component version 4 09:54:11 WARNING:Target server version must have a certified Storage Manager installed after conversion/reversion and before bringing up server. 09:54:11 Conv/rev: Finished conversion for component BAR 09:54:11 Conv/rev: Started conversion for component PLOAD 09:54:11 Conv/rev: Finished conversion for component PLOAD 09:54:11 Conv/rev: Started conversion for component RTREE 09:54:11 *** RTREE 1 9.40xC1 *** 09:55:14 Process exited with return code 1: /bin/sh /bin/sh -c /apps/informix/etc/no_log.sh 3 6 "Internal Subsystem failure: 'SMI'" "Error building 'sysmaster' database." "" 6054 09:59:08 Checkpoint Completed: duration was 0 seconds. 09:59:08 Wed Jan 12 - loguniq 268861, logpos 0x3d33018, timestamp: 0x14d48ad1 Interval: 7 09:59:08 Maximum server connections 0 09:59:08 Checkpoint Statistics - Avg. Txn Block Time 0.000, # Txns blocked 0, Plog used 47, Llog used 2 10:14:17 Checkpoint Completed: duration was 0 seconds. 10:14:17 Wed Jan 12 - loguniq 268861, logpos 0x3d35018, timestamp: 0x14d48e6e Interval: 8 10:14:17 Maximum server connections 0 10:14:17 Checkpoint Statistics - Avg. Txn Block Time 0.000, # Txns blocked 0, Plog used 2, Llog used 2 10:14:54 Conv/rev: Failed to convert component RTREE to component version 2 10:14:54 Conv/rev: Failed to convert for component RTREE 10:14:54 Conversion failed. 10:14:54 *** Conversion failed *** CONVERSION_GUARD was on. To go back to the original version of IDS, run onrestorept. 10:14:54 Assert Failed: *** Conversion failed *** CONVERSION_GUARD was on. To go back to the original version of IDS, run onrestorept. 10:14:54 IBM Informix Dynamic Server Version 11.70.FC1GE 10:14:54 Who: Session(8, informix@pabaanqa, 0, c0000000384cfc90) Thread(147, convert, c00000003848cdc8, 3) File: framework.c Line: 1011 10:14:54 Results: Dynamic Server must abort 10:14:54 Action: Restart conversion after fixing the problem. 10:14:54 stack trace for pid 15002 written to /apps/informix/tmp/af.47bbdaf 10:14:54 See Also: /apps/informix/tmp/af.47bbdaf 10:14:54 framework.c, line 1011, thread 147, proc id 15002, *** Conversion failed *** CONVERSION_GUARD was on. To go back to the original version of IDS, run onrestorept.. 10:14:54 invoke_alarm(): /bin/sh -c '/apps/informix/etc/no_log.sh 5 6 "Internal Subsystem failure: 'MT'" "framework.c, line 1011, thread 147, proc id 15002, *** Conversion failed *** CONVERSION_GUARD was on. To go back to the original version of IDS, run onrestorept.." "" 6500' 10:14:54 invoke_alarm(): mt_exec failed, status 127, errno 0 10:14:54 Fatal error in ADM VP at mt.c:14250 10:14:54 Unexpected virtual processor termination, pid = 15002, exit = 0x100 10:14:54 PANIC: Attempting to bring system down 10:49:46 IBM Informix Dynamic Server Started. 10:49:47 Segment locked: addr=c00000000155e000, size=911343616

yes, but what about See '/tmp/buildsmi.15000'. From: "TRI TRINH" <tri_trinh@hotmail.com> To: ids@iiug.org Date: 01/13/2011 09:33 AM Subject: Upgrade from 10.00.FC6 - 11.70.FC1 [22406] Sent by: ids-bounces@iiug.org We are trying to upgrade our IDS engine in HP-UX B.11.11 from version 10.00.FC6 to version 11.70.FC1. We encountered error during the conversion. Please take a look at the error. Do we need to drop the sysmaster table from the old version before the conversion? Do we need a full backup from the old version for the conversion to run? 09:44:53 Conv/rev: Finished conversion for component SQL to component version 33 09:44:53 The dummy updates succeeded while converting database sysuser. 09:44:53 Internal Conversion Completed Successfuly 09:44:53 Conversion Enabling Client Connections 09:44:53 Conv/rev: Finished conversion for component SQL 09:44:53 Conv/rev: Started conversion for component SMI 09:44:53 *** SMI 1 11.70xC1 *** 09:44:53 Build of sysmaster and sysutils databases Started 09:44:53 Converting/Rebuilding 'sysmaster' database ... 09:44:54 Single-User Mode. 09:45:23 Process exited with return code 1: /bin/sh /bin/sh -c /apps/informix/etc/no_log.sh 2 23 "" "Logical Log 268859 Complete, timestamp: 0x14b87b4d." "" 23001 09:45:43 Process exited with return code 1: /bin/sh /bin/sh -c /apps/informix/etc/no_log.sh 2 23 "" "Logical Log 268860 Complete, timestamp: 0x14cc49bb." "" 23001 09:49:03 Checkpoint Completed: duration was 25 seconds. 09:49:03 Wed Jan 12 - loguniq 268861, logpos 0x3d31018, timestamp: 0x14d48991 Interval: 6 09:49:03 Maximum server connections 0 09:49:03 Checkpoint Statistics - Avg. Txn Block Time 0.000, # Txns blocked 0, Plog used 13312, Llog used 43437 09:54:11 Error building 'sysmaster' database. 09:54:11 See '/tmp/buildsmi.15000'. 09:54:11 Conv/rev: Finished conversion for component SMI to component version 2 09:54:11 Conv/rev: Finished conversion for component SMI 09:54:11 Conv/rev: Started conversion for component BAR 09:54:11 *** BAR 1 11.10xC1 *** 09:54:11 Conv/rev: Finished conversion for component BAR to component version 3 09:54:11 *** BAR 1 11.50xC1 *** 09:54:11 Conv/rev: Finished conversion for component BAR to component version 4 09:54:11 WARNING:Target server version must have a certified Storage Manager installed after conversion/reversion and before bringing up server. 09:54:11 Conv/rev: Finished conversion for component BAR 09:54:11 Conv/rev: Started conversion for component PLOAD 09:54:11 Conv/rev: Finished conversion for component PLOAD 09:54:11 Conv/rev: Started conversion for component RTREE 09:54:11 *** RTREE 1 9.40xC1 *** 09:55:14 Process exited with return code 1: /bin/sh /bin/sh -c /apps/informix/etc/no_log.sh 3 6 "Internal Subsystem failure: 'SMI'" "Error building 'sysmaster' database." "" 6054 09:59:08 Checkpoint Completed: duration was 0 seconds. 09:59:08 Wed Jan 12 - loguniq 268861, logpos 0x3d33018, timestamp: 0x14d48ad1 Interval: 7 09:59:08 Maximum server connections 0 09:59:08 Checkpoint Statistics - Avg. Txn Block Time 0.000, # Txns blocked 0, Plog used 47, Llog used 2 10:14:17 Checkpoint Completed: duration was 0 seconds. 10:14:17 Wed Jan 12 - loguniq 268861, logpos 0x3d35018, timestamp: 0x14d48e6e Interval: 8 10:14:17 Maximum server connections 0 10:14:17 Checkpoint Statistics - Avg. Txn Block Time 0.000, # Txns blocked 0, Plog used 2, Llog used 2 10:14:54 Conv/rev: Failed to convert component RTREE to component version 2 10:14:54 Conv/rev: Failed to convert for component RTREE 10:14:54 Conversion failed. 10:14:54 *** Conversion failed *** CONVERSION_GUARD was on. To go back to the original version of IDS, run onrestorept. 10:14:54 Assert Failed: *** Conversion failed *** CONVERSION_GUARD was on. To go back to the original version of IDS, run onrestorept. 10:14:54 IBM Informix Dynamic Server Version 11.70.FC1GE 10:14:54 Who: Session(8, informix@pabaanqa, 0, c0000000384cfc90) Thread(147, convert, c00000003848cdc8, 3) File: framework.c Line: 1011 10:14:54 Results: Dynamic Server must abort 10:14:54 Action: Restart conversion after fixing the problem. 10:14:54 stack trace for pid 15002 written to /apps/informix/tmp/af.47bbdaf 10:14:54 See Also: /apps/informix/tmp/af.47bbdaf 10:14:54 framework.c, line 1011, thread 147, proc id 15002, *** Conversion failed *** CONVERSION_GUARD was on. To go back to the original version of IDS, run onrestorept.. 10:14:54 invoke_alarm(): /bin/sh -c '/apps/informix/etc/no_log.sh 5 6 "Internal Subsystem failure: 'MT'" "framework.c, line 1011, thread 147, proc id 15002, *** Conversion failed *** CONVERSION_GUARD was on. To go back to the original version of IDS, run onrestorept.." "" 6500' 10:14:54 invoke_alarm(): mt_exec failed, status 127, errno 0 10:14:54 Fatal error in ADM VP at mt.c:14250 10:14:54 Unexpected virtual processor termination, pid = 15002, exit = 0x100 10:14:54 PANIC: Attempting to bring system down 10:49:46 IBM Informix Dynamic Server Started. 10:49:47 Segment locked: addr=c00000000155e000, size=911343616 ******************************************************************************* Forum Note: Use "Reply" to post a response in the discussion forum.

The file '/tmp/buildsmi.15000' is no longer there for some reason. I searched the whole server, but it's not there. You "YES" means we have to drop the sysmaster or we need a full backup?

If we drop the sysmaster, then the revert (CONVERSION_GUARD) won't work. :(.

Did you reboot the system? A lot of systems will clean the /tmp directory upon reboot. From: "TRI TRINH" <tri_trinh@hotmail.com> To: ids@iiug.org Date: 01/13/2011 09:54 AM Subject: Re: Upgrade from 10.00.FC6 - 11.70.FC1 [22409] Sent by: ids-bounces@iiug.org The file '/tmp/buildsmi.15000' is no longer there for some reason. I searched the whole server, but it's not there. You "YES" means we have to drop the sysmaster or we need a full backup? ******************************************************************************* Forum Note: Use "Reply" to post a response in the discussion forum.

No, we did not reboot the server.

I did a second attempt for the conversion, and it failed at the same spot. Please check the link out. It's hard to believe that we have to drop the sysmaster database before the conversion. https://www-304.ibm.com/support/docview.wss?uid=swg21427622

It's more probable that your rootdbs ran out of space and there wasn't enough space to create the new sysmaster database. From: "TRI TRINH" <tri_trinh@hotmail.com> To: ids@iiug.org Date: 01/13/2011 02:05 PM Subject: Re: Upgrade from 10.00.FC6 - 11.70.FC1 [22421] Sent by: ids-bounces@iiug.org I did a second attempt for the conversion, and it failed at the same spot. Please check the link out. It's hard to believe that we have to drop the sysmaster database before the conversion. https://www-304.ibm.com/support/docview.wss?uid=swg21427622 ******************************************************************************* Forum Note: Use "Reply" to post a response in the discussion forum.

This is from our production instance. Our test instance is a dup of the production. All dbspaces have more than 20% or 2MB of free space. Sorry it's hard to read. I seperated the columns by ### Baan _socl ### Name ### Size ### Pages free ### Kfree ### MB Used ### MB Free ### %free 1 ### rootdbs ### 250000 ### 146014 ### 292028 ### 203.1 ### 285.18 ### 58.41 * 2 ### phylog ### 250000 ### 224947 ### 449894 ### 48.93 ### 439.35 ### 89.98 * 3 ### baandbs ### 250000 ### 46872 ### 93744 ### 396.73 ### 91.55 ### 18.75 4 ### logdbs ### 8704000 ### 2043923 ### 4087846 ### 13007.96 ### 3992.04 ### 23.48 * 5 ### tempdbs ### 1536000 ### 1522378 ### 3044756 ### 26.61 ### 2973.39 ### 99.11 * 6 ### table1041 ### 9216000 ### 2308406 ### 4616812 ### 13491.39 ### 4508.61 ### 25.05 7 ### table1042 ### 17920000 ### 6683026 ### 13366052 ### 21947.21 ### 13052.79 ### 37.29 8 ### table199 ### 20480000 ### 6897936 ### 13795872 ### 26527.47 ### 13472.53 ### 33.68 9 ### index1 ### 28398000 ### 12536145 ### 25072290 ### 30980.19 ### 24484.66 ### 44.14 10 ### table203 ### 13414400 ### 2408051 ### 4816102 ### 21496.78 ### 4703.22 ### 17.95 11 ### table204 ### 15360000 ### 4671780 ### 9343560 ### 20875.43 ### 9124.57 ### 30.42 12 ### indexa ### 19456000 ### 7867210 ### 15734420 ### 22634.36 ### 15365.64 ### 40.44 13 ### table200 ### 22528000 ### 8139916 ### 16279832 ### 28101.73 ### 15898.27 ### 36.13 14 ### table201 ### 24576000 ### 8255245 ### 16510490 ### 31876.47 ### 16123.53 ### 33.59 15 ### table202 ### 20480000 ### 8188793 ### 16377586 ### 24006.26 ### 15993.74 ### 39.98 16 ### logdbs2 ### 9216000 ### 2555923 ### 5111846 ### 13007.96 ### 4992.04 ### 27.73 * 17 ### logdbs3 ### 9216000 ### 2555923 ### 5111846 ### 13007.96 ### 4992.04 ### 27.73 * 18 ### int_trans ### 32768000 ### 19292053 ### 38584106 ### 26320.21 ### 37679.79 ### 58.87 19 ### table205 ### 10240000 ### 223742 ### 447484 ### 19563 ### 437 ### 2.18 20 ### table206 ### 12288000 ### 277954 ### 555908 ### 23457.12 ### 542.88 ### 2.26 21 ### table207 ### 12288000 ### 324178 ### 648356 ### 23366.84 ### 633.16 ### 2.64 22 ### table106 ### 1024000 ### 1020119 ### 2040238 ### 7.58 ### 1992.42 ### 99.62 23 ### table107 ### 1024000 ### 1018882 ### 2037764 ### 10 ### 1990 ### 99.5 Not e: * Not include d in tot al 335058.77 176094.36

On Thu, Jan 13, 2011 at 9:14 PM, TRI TRINH <tri_trinh@hotmail.com> wrote: > This is from our production instance. Our test instance is a dup of the > production. All dbspaces have more than 20% or 2MB of free space. Sorry > it's > hard to read. I seperated the columns by ### > > Baan _socl > > ### Name ### Size ### Pages free ### Kfree ### MB Used ### MB Free ### > %free > 1 ### rootdbs ### 250000 ### 146014 ### 292028 ### 203.1 ### 285.18 ### > 58.41 > * > 2 ### phylog ### 250000 ### 224947 ### 449894 ### 48.93 ### 439.35 ### > 89.98 * > 3 ### baandbs ### 250000 ### 46872 ### 93744 ### 396.73 ### 91.55 ### 18.75 > 4 ### logdbs ### 8704000 ### 2043923 ### 4087846 ### 13007.96 ### 3992.04 > ### > 23.48 * > 5 ### tempdbs ### 1536000 ### 1522378 ### 3044756 ### 26.61 ### 2973.39 ### > 99.11 * > 6 ### table1041 ### 9216000 ### 2308406 ### 4616812 ### 13491.39 ### > 4508.61 > ### 25.05 > 7 ### table1042 ### 17920000 ### 6683026 ### 13366052 ### 21947.21 ### > 13052.79 ### 37.29 > 8 ### table199 ### 20480000 ### 6897936 ### 13795872 ### 26527.47 ### > 13472.53 > ### 33.68 > 9 ### index1 ### 28398000 ### 12536145 ### 25072290 ### 30980.19 ### > 24484.66 > ### 44.14 > 10 ### table203 ### 13414400 ### 2408051 ### 4816102 ### 21496.78 ### > 4703.22 > ### 17.95 > 11 ### table204 ### 15360000 ### 4671780 ### 9343560 ### 20875.43 ### > 9124.57 > ### 30.42 > 12 ### indexa ### 19456000 ### 7867210 ### 15734420 ### 22634.36 ### > 15365.64 > ### 40.44 > 13 ### table200 ### 22528000 ### 8139916 ### 16279832 ### 28101.73 ### > 15898.27 ### 36.13 > 14 ### table201 ### 24576000 ### 8255245 ### 16510490 ### 31876.47 ### > 16123.53 ### 33.59 > 15 ### table202 ### 20480000 ### 8188793 ### 16377586 ### 24006.26 ### > 15993.74 ### 39.98 > 16 ### logdbs2 ### 9216000 ### 2555923 ### 5111846 ### 13007.96 ### 4992.04 > ### 27.73 * > 17 ### logdbs3 ### 9216000 ### 2555923 ### 5111846 ### 13007.96 ### 4992.04 > ### 27.73 * > 18 ### int_trans ### 32768000 ### 19292053 ### 38584106 ### 26320.21 ### > 37679.79 ### 58.87 > 19 ### table205 ### 10240000 ### 223742 ### 447484 ### 19563 ### 437 ### > 2.18 > 20 ### table206 ### 12288000 ### 277954 ### 555908 ### 23457.12 ### 542.88 > ### > 2.26 > 21 ### table207 ### 12288000 ### 324178 ### 648356 ### 23366.84 ### 633.16 > ### > 2.64 > 22 ### table106 ### 1024000 ### 1020119 ### 2040238 ### 7.58 ### 1992.42 > ### > 99.62 > 23 ### table107 ### 1024000 ### 1018882 ### 2037764 ### 10 ### 1990 ### > 99.5 > Not e: * Not include d in tot al 335058.77 176094.36 > > > > ******************************************************************************* > Forum Note: Use "Reply" to post a response in the discussion forum. > > Sorry if I'm loosing something, but if this is a Baan system you may need more space than in "normal" cases due to the number of tables that you can get in this kind of system. If I understood you correctly you say you have at least 2MB in each dbspace. This will be too short! I was involved in a similar situation and I had this problem. In any case, you should open a PMR in order to get some proper help. Thanks and regards. -- Fernando Nunes Portugal http://informix-technology.blogspot.com My email works... but I don't check it frequently... --000e0cd1d2d8cfa4000499d271c2

The 20% and 2MB are requirements from ids_mig_bookmap. We have lots more space available than the requirements. We have an open PMR for the problem. Thank you anyhow. tri For a successful conversion of each database, ensure that 2000 KB of free space per database is available in each dbspace where a database resides. To ensure enough free space is available: 1. Calculate the amount of free space that each dbspace requires. In the following equation, n is the number of databases in the dbspace and X is the amount of free space they require: X kilobytes free space = 2000 kilobytes * n The minimum number of databases is 2 (for the sysmaster and sysadmin databases).

We ran the query that recommended from IBM to check space for conversion. The result from the query said we need more space for dbspace 1 and 3. We 8GB each into dbspace 1 and 3, and ran the query again. We got the same result from the query. How do we update the catalog from the sysmaster? Thanks DATABASE sysmaster;SELECT partdbsnum(partnum) dbspace_num, trunc(count(*) * 2000) free_space_req FROM sysdatabases GROUP BY 1 ORDER BY 1; Result from query: dbspace_num free_space_req 1 10000 3 2000

We ran this query and the newly added space show up. SELECT dbsnum dbspace_num, sum(nfree) free_space_avail FROM syschunks GROUP BY 1 ORDER BY 1; Result: dbspace_num free_space_avail 1 4240686 2 224947 3 4142860 4 2043923 5 1522386 6 2308306 7 6682990 8 6897936 9 12540414 10 2408051 11 4671780 12 7867210 13 8139916 14 8255245 15 8188793 16 2555923 17 2555923 18 19292053 19 2271736 20 2325948 21 2372172 22 1020119 23 1018882

Hi, I added more chucks into the dbspaces, and ran the conversion again. Here is the latest message. Any ideas would be appreciated. 11:47:25 Succeeded 11:47:25 Converting Partition Header Page 284:2017. 11:47:25 Converting Partition Header from version 10.00 11:47:25 Succeeded 11:47:26 Onconfig parameter JVPJAVAVM modified from libjava.so to <blank>. 11:47:26 Onconfig parameter JVPLOGFILE modified from jvp.log to /apps/informix/tmp/jvp.log. 11:47:26 Onconfig parameter JVPPROPFILE modified from .jvpprops to /apps/informix/extend/krakatoa/.jvpprops. 11:47:26 Onconfig parameter JVPCLASSPATH modified from <blank> to /apps/informix/extend/krakatoa/krakatoa.jar:/apps/informix/extend/krakatoa/jd bc.jar. 11:47:26 Onconfig parameter LOGBUFF modified from 32 to 64. 11:47:26 Onconfig parameter CKPTINTVL modified from 600 to 300. 11:47:26 Onconfig parameter LTXHWM modified from 45 to 70. 11:47:26 Onconfig parameter LTXEHWM modified from 55 to 80. 11:47:26 Onconfig parameter DUMPDIR modified from /tmp to /apps/informix/tmp. 11:47:26 Onconfig parameter NUMAIOVPS modified from 10 to 2. 11:47:26 Onconfig parameter ALARMPROGRAM modified from /apps/informix/etc/no_log.sh to /apps/informix/etc/log_full.sh. 11:47:26 Onconfig parameter VP_MEMORY_CACHE_KB modified from 0 to 612000. 11:47:26 Onconfig parameter DS_TOTAL_MEMORY modified from 1000000 to 100000. 11:47:26 Onconfig parameter BAR_DEBUG_LOG modified from /apps/informix/tmp/bar_debug.log to /apps/informix/logs_test/bar_act.log. 11:47:26 Dataskip is now OFF for all dbspaces 11:47:26 Physical recovery completed, continuing Conversion 11:47:26 Conv/rev: Started check phase of conversion for component RSAM 11:47:26 *** RSAM 1 11.10xC1 *** 11:47:26 Conv/rev: Finished check phase of conversion for component RSAM to component version 15 11:47:26 Conversion failed for component RSAM : entity_ver 15 != target version 25 11:47:26 Assert Failed: Conv/rev: Failed in check phase of conversion for component RSAM 11:47:26 IBM Informix Dynamic Server Version 11.70.FC1GE 11:47:26 Who: Session(8, informix@pabaanqa, 0, c0000000384d0930) Thread(151, convert, c00000003848f048, 1) File: framework.c Line: 743 11:47:26 Results: Dynamic Server must abort 11:47:26 Action: Restart conversion after fixing the problem. 11:47:26 stack trace for pid 25535 written to /apps/informix/tmp/af.47fc41e 11:47:26 See Also: /apps/informix/tmp/af.47fc41e 11:47:26 framework.c, line 743, thread 151, proc id 25535, Conv/rev: Failed in check phase of conversion for component RSAM. 11:47:26 invoke_alarm(): /bin/sh -c '/apps/informix/etc/log_full.sh 5 6 "Internal Subsystem failure: 'MT'" "framework.c, line 743, thread 151, proc id 25535, Conv/rev: Failed in check phase of conversion for component RSAM." "" 6500' 11:47:26 invoke_alarm(): mt_exec failed, status 127, errno 0 11:47:26 The Master Daemon Died 11:47:27 PANIC: Attempting to bring system down content of af.47fc41e ====================== 11:47:26 11:47:26 IBM Informix Dynamic Server Version 11.70.FC1GE Software Serial Number AAA#B000000 11:47:26 Assert Failed: Conv/rev: Failed in check phase of conversion for component RSAM 11:47:26 Who: Session(8, informix@pabaanqa, 0, c0000000384d0930) Thread(151, convert, c00000003848f048, 1) File: framework.c Line: 743 11:47:26 Results: Dynamic Server must abort 11:47:26 Action: Restart conversion after fixing the problem. 11:47:26 Raw hex dump of stack located in /apps/informix/tmp/af.47fc41e.rawstk 11:47:26 Stack for thread: 151 convert base: 0xc0000000415cb000 len: 69632 pc: 0x0000000000000000 tos: 0xc0000000415cd160 state: running vp: 1 ( 0) 0x4000000001274b08 legacy_hp_afstack + 0x320 [/apps/informix/bin/oninit] ( 1) 0x4000000001273f94 afstack + 0x64 [/apps/informix/bin/oninit] ( 2) 0x4000000001272e3c afhandler + 0xb54 [/apps/informix/bin/oninit] ( 3) 0x40000000012722d8 afcrash_interface + 0x50 [/apps/informix/bin/oninit] ( 4) 0x4000000000deb038 do_each_version_work + 0xdc0 [/apps/informix/bin/oninit] ( 5) 0x4000000000deb238 generic_fw_main + 0x108 [/apps/informix/bin/oninit] ( 6) 0x4000000000debb84 convertSM + 0x5c [/apps/informix/bin/oninit] ( 7) 0x4000000000de9ce4 fwFacade + 0x17c [/apps/informix/bin/oninit] ( 8) 0x4000000000d6bd88 convert + 0x208 [/apps/informix/bin/oninit] ( 9) 0x400000000123f074 startup + 0xd4 [/apps/informix/bin/oninit] (10) 0x400000000123ee5c resume + 0x10c [/apps/informix/bin/oninit] base: 0xc0000000415cb000 len: 69632 pc: 0x0000000000000000 tos: 0xc0000000415cd160 state: running vp: 1 11:47:26 See Also: /apps/informix/tmp/af.47fc41e 11:47:26 sh /apps/informix/etc/evidence.sh 3 0 /apps/informix/tmp/af.47fc41e 8 0xc00000003848f048 151 0xc0000000415ca028 1537 0 0 0 0 11:47:26 ------------------ End of assertion failure 0 -----------------

Hi, I just want to update the issues that we encountered. The ids_ machine_notes_11.70 recommended the kernel parameters: nfile: 5000 semmni: 3000 semmns: 3000 semmnu: 1400 shmmax: 4398046511104 shmmni: 500 shmseg: 500 The conversion generated warning in the /var/adm/syslog with Jan 21 09:48:19 pabaanqa vmunix: file: table is ffuilll Jan 21 09:48:19 pabaanqa vmunix: e: table is full Jan 21 09:48:19 pabaanqa vmunix: filef:i ltea:b leta bilse ifulls The new IDS 11.70 uses lot more Unix processes, so before conversion we increased the two kernel parameters, and “sar –v 1 3600” and ps –ef |grep Informix |wc –l to monitor nfile and maxuprc during the conversion process. Make sure the sysmaster and sysadmin get convert at the end of the conversion process. Nfile => 15,000 maxuprc=300