[Glass] VM temporary object memory is full

Dale Henrichs via Glass glass at lists.gemtalksystems.com
Fri Aug 28 10:16:05 PDT 2015


Bruno,

Looking at what happened when you clear things I would say that the 
following should reinitialize correctly:

[
   WADispatcher default.
   WASystemConfiguration allSubclasses
     do: [ :each | each instance buildDescription ] ]
     ensure: [ System commitTransaction ]

If there are going to be problems, you get to see it during this step 
instead of during lazy initialization ... Then re-register your 
applications and you should be in shape ...

As I write this I go ... okay what state related to 
WASystemConfiguration gets changed by initializers being run during the 
load ... this might be the place where you are having the trouble ... 
when gemstone loads initalizers are run that may change some of the 
critical defaults  ...
oh and I haven't tried the above code myself yet ...

Let me know if you've already done the above and/or the above runs 
without error ... while I try to destroy my own seaside state:)

Dale

On 8/28/15 9:47 AM, Dale Henrichs wrote:
> Bruno,
>
> I have a few spare minutes...
>
> Okay, so this error was dumped to the gem log before running out of 
> memory?
>
> The static handler approach might get better information for the exact 
> point that you are running out of memory, but it would be interesting 
> to understand the InternalError ... there are some ProcessorScheduler 
> issues in older versions of GemStone, depending upon version, so I'm 
> sorta curious what version of GemStone ... also depending upon what 
> version of Zinc you are using an infinite loop (under error 
> conditions) is possible as I recall addressing that particular 
> problem  ...
>
> Buuuut, if you hadn't seen this problem prior to the re-initialize 
> event it is likely that this is a side-effect as well ...
>
> What version of Seaside are you using? I will go back to your original 
> message (if you already told us) and try to track down what gets 
> cleared by the expressions used and work backwards from there (as I 
> find the time:)
>
> Dale
>
>
> On 8/28/15 7:01 AM, BrunoBB via Glass wrote:
>> James,
>>
>> I think all my problems comes from Seaside and it is possible that i 
>> broke
>> Seaside (trying to find out how to restore or initialize Seaside again).
>> I thought that the following did the try but ..:
>> WAEnvironment configureApplicationDefaults;
>>     configureApplicationExceptionHandlingDefaults;
>>     registerDefaultComponents;
>>     registerDefaultRequestHandlers.
>>
>> Searching how to initialize everything again :)
>>
>> After:
>> [WAGsZincAdaptor startOn: 8888]
>>    on: AlmostOutOfMemory enable
>>    do: [:ex | ex error: ex description].
>>
>> I got the following stacj error (after this i got the VM TOM is full):
>> InternalError (AbstractException) >> _signalFromPrimitive: @5 line 15
>> ProcessorScheduler >> _addDelay:to: @1 line 1
>> ProcessorScheduler >> _waitForSocket:timeout: @8 line 18
>> GsSocket >> readWillNotBlockWithin: @22 line 37
>> SocketStreamSocket (SpSocket) >> 
>> waitForNonBlockingReadActivityUpToMs: @3
>> line 6
>> SocketStreamSocket >> waitForDataFor:ifClosed:ifTimedOut: @20 line 14
>> SocketStreamSocket >> waitForDataFor: @2 line 6
>> SocketStreamSocket >> receiveDataSignallingTimeout:into:startingAt: 
>> @2 line
>> 8
>> SocketStream >> receiveData @5 line 12
>> SocketStream >> next @7 line 9
>> ZnLineReader >> processNext @5 line 4
>> ZnLineReader >> nextLine @3 line 3
>> ZnRequestLine >> readFrom: @3 line 3
>> ZnRequestLine class >> readFrom: @3 line 3
>> ZnRequest >> readHeaderFrom: @2 line 2
>> ZnRequest (ZnMessage) >> readBinaryFrom: @2 line 3
>> ZnRequest class (ZnMessage class) >> readBinaryFrom: @3 line 3
>> [] in  ExecBlock1 (ZnZincServerAdaptor) >> 
>> configureServerForBinaryReading
>> @2 line 4
>> [] in  ExecBlock0 (ZnSingleThreadedServer) >> readRequest: @3 line 6
>> [] in  ExecBlock0 (ZnSingleThreadedServer) >> 
>> withMaximumEntitySizeDo: @2
>> line 6
>> [] in  ExecBlock0 (DynamicVariable class) >> value:during: @3 line 9
>> ExecBlock0 (ExecBlock) >> ensure: @2 line 12
>> ZnMaximumEntitySize class (DynamicVariable class) >> value:during: @6 
>> line
>> 10
>> ZnManagingMultiThreadedServer (ZnSingleThreadedServer) >>
>> withMaximumEntitySizeDo: @5 line 5
>> ZnManagingMultiThreadedServer (ZnSingleThreadedServer) >> 
>> readRequest: @2
>> line 6
>> [] in  ExecBlock0 (ZnMultiThreadedServer) >> readRequestSafely: @2 
>> line 5
>> ExecBlock0 (ExecBlock) >> on:do: @3 line 42
>> [] in  ExecBlock0 (ZnMultiThreadedServer) >> readRequestSafely: @3 
>> line 6
>> ExecBlock0 (ExecBlock) >> on:do: @3 line 42
>> ZnManagingMultiThreadedServer (ZnMultiThreadedServer) >> 
>> readRequestSafely:
>> @3 line 13
>> ZnManagingMultiThreadedServer (ZnMultiThreadedServer) >>
>> executeOneRequestResponseOn: @2 line 7
>> [] in  ExecBlock0 (ZnMultiThreadedServer) >> 
>> executeRequestResponseLoopOn:
>> @2 line 10
>> [] in  ExecBlock0 (DynamicVariable class) >> value:during: @3 line 9
>> ExecBlock0 (ExecBlock) >> ensure: @2 line 12
>> ZnCurrentServer class (DynamicVariable class) >> value:during: @6 
>> line 10
>> ZnManagingMultiThreadedServer (ZnMultiThreadedServer) >>
>> executeRequestResponseLoopOn: @4 line 8
>> [] in  ExecBlock0 (ZnMultiThreadedServer) >> serveConnectionsOn: @2 
>> line 12
>> ExecBlock0 (ExecBlock) >> on:do: @3 line 42
>> [] in  ExecBlock0 (ZnMultiThreadedServer) >> serveConnectionsOn: @2 
>> line 13
>> ExecBlock0 (ExecBlock) >> ensure: @2 line 12
>> [] in  ExecBlock0 (ZnMultiThreadedServer) >> serveConnectionsOn: @2 
>> line 15
>> [] in  ExecBlock0 (ExecBlock) >> ifCurtailed: @2 line 6
>> ExecBlock0 (ExecBlock) >> ensure: @2 line 12
>> ExecBlock0 (ExecBlock) >> ifCurtailed: @3 line 8
>> [] in  ExecBlock (ZnMultiThreadedServer) >> serveConnectionsOn: @2 
>> line 18
>> GsProcess >> _start @7 line 16
>> UndefinedObject (GsNMethod class) >> _gsReturnToC @1 line 1
>>
>> Regards,
>> Bruno
>>
>>
>>
>> -- 
>> View this message in context: 
>> http://forum.world.st/VM-temporary-object-memory-is-full-tp4846466p4846658.html
>> Sent from the GLASS mailing list archive at Nabble.com.
>> _______________________________________________
>> Glass mailing list
>> Glass at lists.gemtalksystems.com
>> http://lists.gemtalksystems.com/mailman/listinfo/glass
>



More information about the Glass mailing list