ids 11.70.TC1IE limits

This question involves IBM Informix Dynamic Server Version 11.70.TC1IE running on Windows Server 2008 Enterprise, service pack 2. This server uses a new Intel I5 processor with 4gb of ram. It seems that IDS freezes several times a day and no programs can get a connection. Even dbaccess freezes if you try to use it. It was my understanding that Innovator-C Edition did not limit connections. It does, however, limit memory to 2097152 KB . What happens when the memory is exceeded? Does it start dropping connections or refusing them or both? I get stuff like this in the log: 14:13:38 IBM Informix Dynamic Server Started. 14:13:39 Warning: The IBM IDS Innovator-C Edition license restriction limits 14:13:39 the total shared memory size for this server to 2097152 KB. 14:13:39 The maximum allowable shared memory was reset to this size to start the database server. 14:13:39 Requested shared memory segment size rounded from 32656KB to 32704KB 14:55:33 Descriptor 1128 has event 0x1, no longer open 14:55:59 Dynamically allocated new virtual shared memory segment (size 8192KB) 14:55:59 Memory sizes:resident:46272 KB, virtual:73664 KB, SHMTOTAL:2097152 KB 14:56:47 Descriptor 1140 has event 0x1, no longer open 14:58:16 Descriptor 1124 has event 0x1, no longer open 14:58:17 Descriptor 1168 has event 0x1, no longer open 14:58:19 Descriptor 1148 has event 0x1, no longer open 14:58:20 Descriptor 1180 has event 0x1, no longer open 14:58:23 Descriptor 1172 has event 0x1, no longer open 14:58:23 Descriptor 1192 has event 0x1, no longer open 14:58:23 Descriptor 1188 has event 0x1, no longer open 14:58:23 Descriptor 1200 has event 0x1, no longer open 14:58:23 Descriptor 1196 has event 0x1, no longer open 14:58:24 Descriptor 1208 has event 0x1, no longer open 14:58:26 Descriptor 1184 has event 0x1, no longer open 14:58:27 Descriptor 1220 has event 0x1, no longer open 14:59:39 Checkpoint Completed: duration was 0 seconds. 14:59:39 Thu Jul 14 - loguniq 95, logpos 0x2c53018, timestamp: 0xb91ddfb Interval: 10290 What is the “event 0x1” message all about? Here are some of the places that we had to restart IDS so far today: ** booted ** 07:54:18 Memory sizes:resident:46272 KB, virtual:40896 KB, SHMTOTAL:2097152 KB 08:11:16 Memory sizes:resident:46272 KB, virtual:49088 KB, SHMTOTAL:2097152 KB 08:25:20 Memory sizes:resident:46272 KB, virtual:57280 KB, SHMTOTAL:2097152 KB 08:33:13 Memory sizes:resident:46272 KB, virtual:65472 KB, SHMTOTAL:2097152 KB 08:41:33 Memory sizes:resident:46272 KB, virtual:73664 KB, SHMTOTAL:2097152 KB 08:52:48 Memory sizes:resident:46272 KB, virtual:81856 KB, SHMTOTAL:2097152 KB ** booted ** 09:11:06 Memory sizes:resident:46272 KB, virtual:40896 KB, SHMTOTAL:2097152 KB 09:14:58 Memory sizes:resident:46272 KB, virtual:49088 KB, SHMTOTAL:2097152 KB 09:18:26 Memory sizes:resident:46272 KB, virtual:57280 KB, SHMTOTAL:2097152 KB 09:26:47 Memory sizes:resident:46272 KB, virtual:65472 KB, SHMTOTAL:2097152 KB 09:29:43 Memory sizes:resident:46272 KB, virtual:73664 KB, SHMTOTAL:2097152 KB 09:45:34 Memory sizes:resident:46272 KB, virtual:81856 KB, SHMTOTAL:2097152 KB 09:52:45 Memory sizes:resident:46272 KB, virtual:90048 KB, SHMTOTAL:2097152 KB 10:01:01 Memory sizes:resident:46272 KB, virtual:98240 KB, SHMTOTAL:2097152 KB 10:29:40 Memory sizes:resident:46272 KB, virtual:106432 KB, SHMTOTAL:2097152 KB 10:47:54 Memory sizes:resident:46272 KB, virtual:114624 KB, SHMTOTAL:2097152 KB 11:02:40 Memory sizes:resident:46272 KB, virtual:122816 KB, SHMTOTAL:2097152 KB ** booted ** 11:41:10 Memory sizes:resident:46272 KB, virtual:40896 KB, SHMTOTAL:2097152 KB 11:52:22 Memory sizes:resident:46272 KB, virtual:49088 KB, SHMTOTAL:2097152 KB 12:01:16 Memory sizes:resident:46272 KB, virtual:57280 KB, SHMTOTAL:2097152 KB 12:05:33 Memory sizes:resident:46272 KB, virtual:65472 KB, SHMTOTAL:2097152 KB 12:16:51 Memory sizes:resident:46272 KB, virtual:73664 KB, SHMTOTAL:2097152 KB ** booted ** 13:05:23 Memory sizes:resident:46272 KB, virtual:40896 KB, SHMTOTAL:2097152 KB As you can see, the SHMTOTAL stays the same (as we should expect since this is the *free* version). Is this squeezing out connections? How can I find out? This client was running IDS9.4 on a dual Zeon machine running Windows 2000 until the the mobo died. They could not find a replacement mobo or a system that would boot the win2k disk. They bought this new server but we could not get ids9.4 to be happy on it so we installed the new IDS 11.7 . I don’t think they will want to give IBM any money for a “paid-for” edition since this one is so unstable. Does anyone think the instability is something other than a “feature” that IBM placed there to encourage purchase of a different version?

Hi Bill. There is more than one issue here, but mostly you need to work out what the memory leak is. Can you watch "onstat -g ses" and "onstat -g mem" to spot which session or shared memory pool is growing? If it's a session, use "onstat -g ses session-id" to find out the details, typically whether it's "ralloc" or "sql". Once we know that, we can advise which of a number of known scenarios might be the problem. Regards, Doug Lawry