Can anyone confirm this new behaviour we've been seeing since we've patched our
servers wioth Service Pack 1 for Windows 2003? The general system performance
is good and stable, but we have one page wich gathers over 25.000 user profiles
to match them with an other table. This page have always run fine, but since
the update, trouble started. We're running CFMX on Jrun4 using JVM version,
1.4.2_08-b03 and the latest Jrun updaters installed. The page stops processing
at 10 percent of the job done with following error:

Exception in thread "CompilerThread0" java.lang.OutOfMemoryError: requested
988600 bytes for Chunk::new. Out of swap space?

We had some major issues earlier this year with performance and memory and had
the help of a macromedia support engineer debugging it, but now problems are
back. Server runs with 1408 megs of heap memory, and at the moment the process
quits, there's only 200 megs used. Here the last pieces of the the log file, we
have the verbose output of the garbage collector on:

28.214: [GC 85513K->74939K(102400K), 0.0089651 secs]
228.472: [GC 86139K->73955K(102016K), 0.0335861 secs]
06/22 16:48:13 metrics Web threads (busy/total): 3/36 Sessions: 95 Total
Memory=102016 Free=28057 Total JDBC timeouts: {jdbc.timeouts} pool size:
{jdbc.poolSize} active connections: {jdbc.activeConnections} query duration:
{jdbc.queryDuration} query count: {jdbc.queryCount}
228.774: [GC 86179K->73491K(104960K), 0.0117079 secs]
229.019: [GC 85715K->73560K(104960K), 0.0088056 secs]
229.327: [GC 89176K->73848K(105088K), 0.0082491 secs]
229.638: [GC 89208K->74003K(104576K), 0.0090903 secs]
229.964: [GC 89683K->75415K(104512K), 0.0085211 secs]
230.274: [GC 89303K->74127K(104064K), 0.0085338 secs]
230.590: [GC 89935K->76535K(104192K), 0.0104464 secs]
230.884: [GC 89463K->74301K(103680K), 0.0095260 secs]
231.240: [GC 90301K->77512K(103616K), 0.0104667 secs]
231.508: [GC 90056K->74179K(103296K), 0.0191976 secs]
231.928: [GC 90435K->76506K(103296K), 0.0091466 secs]
232.358: [GC 90330K->74506K(103232K), 0.0111805 secs]
233.309: [GC 91018K->75463K(104768K), 0.0923728 secs]
06/22 16:48:18 metrics Web threads (busy/total): 2/36 Sessions: 98 Total
Memory=104768 Free=27946 Total JDBC timeouts: {jdbc.timeouts} pool size:
{jdbc.poolSize} active connections: {jdbc.activeConnections} query duration:
{jdbc.queryDuration} query count: {jdbc.queryCount}
234.756: [GC 91975K->75755K(105920K), 0.0263240 secs]
06/22 16:48:23 metrics Web threads (busy/total): 1/36 Sessions: 99 Total
Memory=105920 Free=27324 Total JDBC timeouts: {jdbc.timeouts} pool size:
{jdbc.poolSize} active connections: {jdbc.activeConnections} query duration:
{jdbc.queryDuration} query count: {jdbc.queryCount}
242.284: [GC 94315K->87065K(106240K), 0.0100463 secs]
06/22 16:48:28 metrics Web threads (busy/total): 2/36 Sessions: 100 Total
Memory=106240 Free=12213 Total JDBC timeouts: {jdbc.timeouts} pool size:
{jdbc.poolSize} active connections: {jdbc.activeConnections} query duration:
{jdbc.queryDuration} query count: {jdbc.queryCount}
243.545: [GC 95385K->77080K(105856K), 0.0198786 secs]

As you can see the Full and partial Garbage Collection times does not
increases extensively.

After this line, the out of memory occures and the service crashes. The page
displays the famous Jrun Closed Connection error, which is logical beacuse of
the restarting service. Has anyone seen similar memory related problems since
patching with service pack 1? We've run the procedure on two boxes, the same
error occurs.