Intellij v.7.0.2 Mac OS X Leopard unusable - help needed

Symptoms:

Start IDEA, open default project, work with a few files, starts eating up 100% of both CPU's. Look at activity monitor, see syslog problem, turn off syslog with launchctl. Keep working for a 10-1hr. CPU is eaten up and code inspections take a long time (over 3 mins or more) to complete. Intellij then locks up and I get the "spinning beach ball". Have to restart mac.

Environment:

Java Version 1.5.0_13-119, mixed mode, sharing
IDEA 7.0.2

There are no cache files in the .IntelliJ directories...

Any help is appreciated.

Best,

-Brian

24 comments

Symptoms:

Start IDEA, open default project, work with a few
files, starts eating up 100% of both CPU's. Look at
activity monitor, see syslog problem, turn off syslog
with launchctl. Keep working for a 10-1hr. CPU is
eaten up and code inspections take a long time (over
3 mins or more) to complete. Intellij then locks up
and I get the "spinning beach ball". Have to restart
mac.

Environment:

Java Version 1.5.0_13-119, mixed mode, sharing
IDEA 7.0.2

There are no cache files in the .IntelliJ
directories...

Any help is appreciated.

Best,

-Brian


I can say I don't have any of these problems running 7.0.2 under OS X 10.5.1. The cache files should be located in ~/Library/Caches/IntelliJIDEA7 on OS X.

--
Matthew

0

Thanks for the location information. Did you do a archive/install on of your mac or vanilla upgrade? Wondering if there's an issue with my Java...

Thanks,

-Brian

0

I am having the same issues with 7.0.2 and Leopard. I have been monitoring this closely and have noticed that the root launchd process starts to consume memory the first time IDEA slows down and syslogd follows not too far behind.

Eventually both processes are consuming hundreds of MBs (I've seen both take up to 1.5 GB) and IDEA becomes completely unresponsive. Killing the rogue syslogd process does not help much as the launchd process cannot be killed and I eventually have to restart the machine.

This happens every time I use IntelliJ.

0

Just to add additional information to this thread I use 7.0.2 on Mac OS X 10.4 and have no issues whatsoever.

0

Just to add additional information to this thread I
use 7.0.2 on Mac OS X 10.4 and have no issues
whatsoever.


Thanks for adding to this thread. I believe the issue is with 7.0.2 and Leopard (10.5.X).

Has anyone else experienced issues with this combination? Another random thing that I noticed is that when the "Project" or "Structure" pane is visible and you close active editor windows, you eventually end up with an editor window that spans the entire editor tab pane. This usually causes the launchd and/or syslogd processes to freak out as described above. Switching the "Project" or "Structure" views on/off results in multiple windows showing up in the tab pane again (until you close a few more).

Can someone at IntelliJ offer some advice about this? Is there a way to turn off syslogd logging in IDEA?

0

Is it the same problem as described here? http://intellij.net/forums/thread.jspa?messageID=5201875&#5201875

In that case I have had this problem for a while, but have still not found solution to the problem.

André

0

Is it the same problem as described here?
http://intellij.net/forums/thread.jspa?messageID=52018
75??

In that case I have had this problem for a while, but
have still not found solution to the problem.

André


It sounds like the syslog symptoms are similar, but the underlying problem may not be. My system.log file contains errors like this (which seem to correlate with the strange editor tab pane issues I described above):

0

Thanks for the location information. Did you do a
archive/install on of your mac or vanilla upgrade?
Wondering if there's an issue with my Java...

Thanks,

-Brian


I just did a default upgrade from 10.4.11.

--
Matthew

0

I am having same issues and have been working with IntelliJ Support to resolve.. though no progress so far, and they have not responded to latest emails at all. The system is becoming unusable as it grinds to a halt after a short while of using it.
I have also upgraded to the 7.0 EAP version (Build #7626) with no better result.

This is a REAL problem, and there are enough people seeing it !

The CPU is running wild while error messages are streamed from IDEA to system.log. I have had it as much as 20 minutes backlog is streaming messages to this log.
IntelliJ support had me reconfigure the log.xml file in order to stop the log messages. But this had no effect (and does not address the source of the problem anyway.)
I have pasted below, the specific error that I see streaming to the log literally 100'000's of times. It is an assertTrue coming from intelij code.

Another clue.. i have notices this happening particularly when I opened a web project with jsp's (as opposed to a java app project).

Hopefully the guys at IntelliJ are takiing this seriously and working on it. This has been a fantastic product for years... I would hate to have to abandon it now.

Most common error in /var/log/system.log:
Jan 21 11:10:01 Macintosh .com.jetbrains.intellij[764]: java.lang.Throwable
Jan 21 11:10:01 Macintosh .com.jetbrains.intellij[764]: at com.intellij.openapi.diagnostic.Logger.assertTrue(Logger.java:89)
Jan 21 11:10:01 Macintosh .com.jetbrains.intellij[764]: at com.intellij.psi.impl.file.impl.FileManagerImpl.a(FileManagerImpl.java:13)
Jan 21 11:10:01 Macintosh .com.jetbrains.intellij[764]: at com.intellij.psi.impl.file.impl.FileManagerImpl.findClass(FileManagerImpl.java:90)
Jan 21 11:10:01 Macintosh .com.jetbrains.intellij[764]: at com.intellij.psi.impl.PsiManagerImpl$PsiElementFinderImpl.findClass(PsiManagerImpl.java:22)
Jan 21 11:10:01 Macintosh .com.jetbrains.intellij[764]: at com.intellij.psi.impl.PsiManagerImpl.findClass(PsiManagerImpl.java:297)
Jan 21 11:10:01 Macintosh .com.jetbrains.intellij[764]: at com.intellij.execution.junit.JUnitUtil.b(JUnitUtil.java:42)
Jan 21 11:10:01 Macintosh .com.jetbrains.intellij[764]: at com.intellij.execution.junit.JUnitUtil.a(JUnitUtil.java:67)
Jan 21 11:10:01 Macintosh .com.jetbrains.intellij[764]: at com.intellij.execution.junit.JUnitUtil.a(JUnitUtil.java:60)
Jan 21 11:10:01 Macintosh .com.jetbrains.intellij[764]: at com.intellij.execution.junit.JUnitUtil.isTestClass(JUnitUtil.java:70)
Jan 21 11:10:01 Macintosh .com.jetbrains.intellij[764]: at com.intellij.execution.junit.JUnitTestFramework.isTestKlass(JUnitTestFramework.java:29)
Jan 21 11:10:01 Macintosh .com.jetbrains.intellij[764]: at com.intellij.codeInsight.TestUtil.isTestClass(TestUtil.java:38)
Jan 21 11:10:01 Macintosh .com.jetbrains.intellij[764]: at com.intellij.psi.impl.ElementBase.a(ElementBase.java:117)
Jan 21 11:10:01 Macintosh .com.jetbrains.intellij[764]: at com.intellij.psi.impl.ElementBase.access$000(ElementBase.java:23)
Jan 21 11:10:01 Macintosh .com.jetbrains.intellij[764]: at com.intellij.psi.impl.ElementBase$1.compute(ElementBase.java:1)
.
.
.

0

I got a few colleagues that currently uses Eclipse, but said they are tempted to give IntelliJ a try. I have not said anything yet but if this problem is not addressed by JetBrains I have to tell them not to do the switch.

Please JetBrains, give us some feedback on this one.

André

0

I see the same issue, my syslog is full of completely different stacks than the ones that have been posted. I tried to delete my IDEA caches, but they do not exist under ~/Library/Caches/ at all. Where else could they be?

I have had to simply stop syslogd to make my system usable. Here is the stack from my system.log

Jan 30 12:42:51 Mac .com.jetbrains.intellij[17588]: at sun.reflect.DelegatingMethodAccessorImpl.invoke(DelegatingMethodAccessorImpl.java:25)
Jan 30 12:42:51 Mac .com.jetbrains.intellij[17588]: at java.lang.reflect.Method.invoke(Method.java:585)
Jan 30 12:42:51 Mac .com.jetbrains.intellij[17588]: at com.intellij.util.xml.DomReflectionUtil.invokeMethod(DomReflectionUtil.java:95)
Jan 30 12:42:51 Mac .com.jetbrains.intellij[17588]: at com.intellij.util.xml.impl.VisitorDescription.acceptElement(VisitorDescription.java:13)
Jan 30 12:42:51 Mac .com.jetbrains.intellij[17588]: at com.intellij.util.xml.impl.DomInvocationHandler.accept(DomInvocationHandler.java:315)
Jan 30 12:42:51 Mac .com.jetbrains.intellij[17588]: at sun.reflect.GeneratedMethodAccessor100.invoke(Unknown Source)
Jan 30 12:42:51 Mac .com.jetbrains.intellij[17588]: at sun.reflect.DelegatingMethodAccessorImpl.invoke(DelegatingMethodAccessorImpl.java:25)
Jan 30 12:42:51 Mac .com.jetbrains.intellij[17588]: at java.lang.reflect.Method.invoke(Method.java:585)
Jan 30 12:42:51 Mac .com.jetbrains.intellij[17588]: at com.intellij.util.xml.impl.InvocationCache$6.invoke(InvocationCache.java:1)
Jan 30 12:42:51 Mac .com.jetbrains.intellij[17588]: at com.intellij.util.xml.impl.DomInvocationHandler.a(DomInvocationHandler.java:151)
Jan 30 12:42:51 Mac .com.jetbrains.intellij[17588]: at com.intellij.util.xml.impl.DomInvocationHandler.invoke(DomInvocationHandler.java:185)
Jan 30 12:42:51 Mac .com.jetbrains.intellij[17588]: at com.intellij.util.xml.GenericAttributeValue$$EnhancerByCGLIB$$ebdf6241.accept(]]>)
Jan 30 12:42:51 Mac .com.jetbrains.intellij[17588]: at com.intellij.util.xml.highlighting.DomElementsInspection.checkChildren(DomElementsInspection.java:78)
Jan 30 12:42:51 Mac .com.jetbrains.intellij[17588]: at com.intellij.util.xml.highlighting.DomElementsInspection$1.visitDomElement(DomElementsInspection.java:66)
Jan 30 12:42:51 Mac .com.jetbrains.intellij[17588]: at sun.reflect.GeneratedMethodAccessor102.invoke(Unknown Source)
Jan 30 12:42:51 Mac .com.jetbrains.intellij[17588]: at sun.reflect.DelegatingMethodAccessorImpl.invoke(DelegatingMethodAccessorImpl.java:25)
Jan 30 12:42:51 Mac .com.jetbrains.intellij[17588]: at java.lang.reflect.Method.invoke(Method.java:585)
Jan 30 12:42:51 Mac .com.jetbrains.intellij[17588]: at com.intellij.util.xml.DomReflectionUtil.invokeMethod(DomReflectionUtil.java:95)
Jan 30 12:42:51 Mac .com.jetbrains.intellij[17588]: at com.intellij.util.xml.impl.VisitorDescription.acceptElement(VisitorDescription.java:13)
Jan 30 12:42:51 Mac .com.jetbrains.intellij[17588]: at com.intellij.util.xml.impl.DomInvocationHandler.accept(DomInvocationHandler.java:315)

0

Could you provide CPU snapshot?

0

Sure, here you go. I had to wait for it to happen again. If you constantly stop syslogd it seemed to pass. But soon as I started actively using IDEA again it happened.

Sampling process 19121 for 3 seconds with 1 millisecond of run time between samples
Sampling completed, processing symbols...
Analysis of sampling syslogd (pid 19121) every 1 millisecond
Call graph:
2281 Thread_2503
2281 start
2281 main
2281 klog_in_close
2281 db_archive
2280 asl_store_compact
1334 fseek
1294 _fseeko
722 __sflush
712 write$NOCANCEL$UNIX2003
712 write$NOCANCEL$UNIX2003
8 _swrite
4 _swrite
2 __swrite
2 __swrite
2 write$NOCANCEL$UNIX2003
2 write$NOCANCEL$UNIX2003
2 __sflush
533 lseek
533 lseek
15 _sseek
9 _sseek
4 lseek
4 lseek
2 __error
2 __error
13 __bzero
13 __bzero
7 _fseeko
4 __error
4 __error
12 flockfile
6 __spin_lock
6 __spin_lock
5 pthread_mutex_lock
5 pthread_mutex_lock
1 flockfile
9 pthread_mutex_unlock
9 pthread_mutex_unlock
7 fseek
3 __spin_lock
3 __spin_lock
3 funlockfile
2 funlockfile
1 pthread_self
1 pthread_self
3 memset
3 memset
1 dyld_stub__spin_lock
1 dyld_stub__spin_lock
1 dyld_stub__spin_unlock
1 dyld_stub__spin_unlock
1 spin_lock
1 spin_lock
885 fread
842 __srefill
819 read$NOCANCEL$UNIX2003
819 read$NOCANCEL$UNIX2003
10 __srefill
9 _sread
9 _sread
4 __sflush
4 __sflush
12 flockfile
8 pthread_mutex_lock
8 pthread_mutex_lock
3 __spin_lock
3 __spin_lock
1 flockfile
9 fread
7 __memcpy
7 __memcpy
7 funlockfile
5 pthread_self
5 pthread_self
2 funlockfile
4 pthread_mutex_unlock
4 pthread_mutex_unlock
2 __spin_lock
2 __spin_lock
1 spin_lock
1 spin_lock
1 spin_unlock
1 spin_unlock
30 fwrite$UNIX2003
21 __sfvwrite
11 __memcpy
11 __memcpy
5 __sfvwrite
5 __swsetup
5 __swsetup
3 flockfile
3 __spin_lock
3 __spin_lock
3 fwrite$UNIX2003
1 __spin_lock
1 __spin_lock
1 funlockfile
1 funlockfile
1 pthread_mutex_unlock
1 pthread_mutex_unlock
22 asl_store_compact
5 memcpy
5 memcpy
2 asl_store_error
2 asl_store_error
1 dyld_stub___error
1 dyld_stub___error
1 dyld_stub_memcpy
1 dyld_stub_memcpy
1 dyld_stub_fseek
1 dyld_stub_fseek
2281 Thread_2603
2281 thread_start
2281 pthreadstart
2281 database_server
2281 mach_msg
2281 mach_msg_trap
2281 mach_msg_trap
2281 Thread_2703
2281 thread_start
2281 pthreadstart
2281 db_worker
2281 db_dequeue
2281 pthread_cond_wait$UNIX2003
2281 __semwait_signal
2281 __semwait_signal
2281 Thread_2803
2281 thread_start
2281 pthreadstart
2281 launchd_drain
2281 vprocmgrlog_drain
2281 vproc_mig_log_drain
2281 mach_msg
2281 mach_msg_trap
2281 mach_msg_trap

Total number in stack (recursive counted multiple, when >=5):
6 __spin_lock

Sort by top of stack, same collapsed (when >= 5):
mach_msg_trap 4562
__semwait_signal 2281
read$NOCANCEL$UNIX2003 819
write$NOCANCEL$UNIX2003 714
lseek 537
asl_store_compact 22
__memcpy 18
__spin_lock 18
pthread_mutex_unlock 14
__bzero 13
pthread_mutex_lock 13
__srefill 10
_sread 9
_sseek 9
fread 9
_fseeko 7
fseek 7
__error 6
__sflush 6
pthread_self 6
__sfvwrite 5
__swsetup 5
funlockfile 5
memcpy 5
Sample analysis of process 19121 written to file /dev/stdout

0

JetBrains,
I got the same problem.IDEA hung with launchd/syslogd locked up.
Can't get rid of the rogue IDEA process.

I believe it is Leopard's bug.
Please work with Apple to resolve this issue, it is really unusable as the OP said.

0

Hello stanley,

I got the same problem.IDEA hung with launchd/syslogd locked up.
Can't get rid of the rogue IDEA process.
I believe it is Leopard's bug. Please work with Apple to resolve this
issue, it is really unusable as the OP said.


The recent Selena 7.0.3 EAP builds no longer cause these problems. You can
download them from:
http://www.jetbrains.net/confluence/display/IDEADEV/Selena+EAP

--
Dmitry Jemerov
Development Lead
JetBrains, Inc.
http://www.jetbrains.com/
"Develop with Pleasure!"


0

I have downloaded 7.03, but am still gettting an entire core on mult-core CPU devoted to syslogd
There are 2 differences in 7.0.3
1) Nothing is being written to system.log anyore
2) syslogd stops consuming resources when intellij is killed
Still the same: syslogd consumes an entire cpu when IDEA is runnig

0

I'm having exactly the same problem, and 7.0.3 doesn't fix it. First I noticed the tab view over the editors expanded to fill the width above the editor, and the arrows at the end disappeared. Then I couldn't change files and eventually the app hung.

I'm happy to provide more debug output - I need my IDEA back :(

0

I am still having this issue as well. IDEA is basically unusable. It has gotten better, but it still happens if I come back to IDEA after a period in another application or after typing for a few minutes. I will provide a debug too, just need to enable the snapshots again. Although every time I tried to get a snapshot with the last build, it crashed IDEA.

-Brandon

0

Is there somewhere i can post the cpu snapshot, IDEA is too slow, I can't get the snapshot under 2mb.

thanks.
-Brandon

0

I tried uploading to ftp://ftp.intellij.net/.uploads but was not able to write any files.
-Brandon

0

I am using 7.0.2 and this is what happens in my site:

The tabs above the editor is buggy.

Suppose I opened 15 files and only 10 file tabs are shown due to
the constraint of the width of the screen.
When I left-clicked the end arrows trying to traverse files,
the tabs above the editor did not change accordingly!
It turns out that the editor is showing a file whose tab is not above
the editor. This isn't right.
If I left-clicked the end arrows but hold it longer to have the file list
shown, the tabs go right when I select files from the list.

Then, a few times when I right-clicked a tab and select "Close All ... ",
IDEA hangs. Syslogd locks CPU, continuously logging:

.com.jetbrains.intellij[97798] ERROR - com.intellij.ide.IdeEventQueue - IntelliJ IDEA 7.0.2 Build #7590
.com.jetbrains.intellij[97798] ERROR - com.intellij.ide.IdeEventQueue - JDK: 1.5.0_13
.com.jetbrains.intellij[97798] ERROR - com.intellij.ide.IdeEventQueue - VM: Java HotSpot(TM) Client VM
.com.jetbrains.intellij[97798] ERROR - com.intellij.ide.IdeEventQueue - Vendor: Apple Inc.
.com.jetbrains.intellij[97798] ERROR - com.intellij.ide.IdeEventQueue - OS: Mac OS X
.com.jetbrains.intellij[97798] ERROR - com.intellij.ide.IdeEventQueue - Last Action: CloseAllUnpinnedEditors
.com.jetbrains.intellij[97798] java.lang.ArrayIndexOutOfBoundsException: 8
...

I installed Leopard by upgrade not clean and reinstall.
I did not know if this is the cause of all the problems occurred to me but
not other people.

0

This could actually be related to JetGroovy. I was doing some work in just a java project and IDEA was fine.

How do I upload a really large cpu snapshot?

Thanks.
-Brandon

0

I am just on the way updating to 7.0.3 RC2 EAP. The problem with the CPU consumption is that syslogd scan asl.db and this log database gets really big when doing anything with IDEA. IDEA is logging like hell to system.log and this most likely ends in asl.db in /var/log.

So basically its an error condition in IDEA that is only there with Leopard. This error condition most likely logs way too much things, even though you dont have problems working with the IDE, i mean functionally, of course its not very responsive because of the syslogd process. Thats strange isnt it.

Hopefully this is solved with 7.0.3, because in fact working is not really possible. Of course you can delete /var/log/asl.db every 5 minutes and restart syslogd, but thats a pain. You can even limit asl.db to get so big, but this would be the wrong place to tune things.

Marc

0

I have this problem to with leapord and 7.0.2. I went back to 7.0.1 and it got better. Netbeans seems to be the only IDE that works great with leopard. Now if netbeans could only get a testng plugin.

0

Please sign in to leave a comment.