You are viewing a plain text version of this content. The canonical link for it is here.
Posted to fx-dev@ws.apache.org by "Damitha Kumarage (JIRA)" <ji...@apache.org> on 2007/02/28 09:28:05 UTC
[jira] Created: (SANDESHA2C-27) Do memory leak fixes in Sandesha2c
Do memory leak fixes in Sandesha2c
----------------------------------
Key: SANDESHA2C-27
URL: https://issues.apache.org/jira/browse/SANDESHA2C-27
Project: Sandesha2/C
Issue Type: Bug
Reporter: Damitha Kumarage
Do memory leak fixes in Sandesha2
--
This message is automatically generated by JIRA.
-
You can reply to this email to add a comment to the issue online.
---------------------------------------------------------------------
To unsubscribe, e-mail: sandesha-dev-unsubscribe@ws.apache.org
For additional commands, e-mail: sandesha-dev-help@ws.apache.org
[jira] Commented: (SANDESHA2C-27) Do memory leak fixes in
Sandesha2c
Posted by "Damitha Kumarage (JIRA)" <ji...@apache.org>.
[ https://issues.apache.org/jira/browse/SANDESHA2C-27?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=12572521#action_12572521 ]
Damitha Kumarage commented on SANDESHA2C-27:
--------------------------------------------
Much of the major memory leaks are fixed by now. However there is still work to be done in this area
> Do memory leak fixes in Sandesha2c
> ----------------------------------
>
> Key: SANDESHA2C-27
> URL: https://issues.apache.org/jira/browse/SANDESHA2C-27
> Project: Sandesha2/C
> Issue Type: Bug
> Reporter: Damitha Kumarage
>
> Do memory leak fixes in Sandesha2
--
This message is automatically generated by JIRA.
-
You can reply to this email to add a comment to the issue online.
---------------------------------------------------------------------
To unsubscribe, e-mail: sandesha-dev-unsubscribe@ws.apache.org
For additional commands, e-mail: sandesha-dev-help@ws.apache.org
[jira] Commented: (SANDESHA2C-27) Do memory leak fixes in
Sandesha2c
Posted by "Damitha Kumarage (JIRA)" <ji...@apache.org>.
[ https://issues.apache.org/jira/browse/SANDESHA2C-27?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=12586161#action_12586161 ]
Damitha Kumarage commented on SANDESHA2C-27:
--------------------------------------------
Also we need to be fully aware where and how Axis2/C free msg_ctx and op_ctx structures and soap envelopes.
- In the client application side when operation client is freed operation context is freed. when operatoin context is freed both of it's in and out message contexts are freed. It should be noted that when consecutive application messages are sent using same service client, new operation clients are created after freeing the previous operation client. This has interesting consequences for rm_ping samples because while application message context is held by Sandesha2/C until CreateSequenceResponse message is completed, operation client is freed for that application message and so is op_ctx, if the client send a second message before CreateSequenceResponse message come. If client send the second rm_ping message with adequate pause then this should not happen.
But we cannot expect so from the client. So my suggestion is when we store the application message context we should mark operation context as in-use by callled axis2_op_ctx_set_in_use() function. Currently for all mep cases ref is incremented for op_ctx which is not good. We shoud mard it as set_in_use for only one way case. At sandesha2_storage_remove_msg_ctx() function we can unset this and free it.
Note that sandesha2_storage_remove_msg_ctx() is called for held application message context when CreateSequenceResponse message complete so in this function we can extract the op_ctx and free it.
- In server side message contexts and opeation contexts are freed at the end of transport worker's process_request function. Note that there, before freeing operation context, it's message contexts are freed.
- Soap envelopes are freed when message context is freed in Axis2/C so we need not specifically worrry about this.
> Do memory leak fixes in Sandesha2c
> ----------------------------------
>
> Key: SANDESHA2C-27
> URL: https://issues.apache.org/jira/browse/SANDESHA2C-27
> Project: Sandesha2/C
> Issue Type: Bug
> Reporter: Damitha Kumarage
>
> Do memory leak fixes in Sandesha2
--
This message is automatically generated by JIRA.
-
You can reply to this email to add a comment to the issue online.
---------------------------------------------------------------------
To unsubscribe, e-mail: sandesha-dev-unsubscribe@ws.apache.org
For additional commands, e-mail: sandesha-dev-help@ws.apache.org
[jira] Commented: (SANDESHA2C-27) Do memory leak fixes in
Sandesha2c
Posted by "Damitha Kumarage (JIRA)" <ji...@apache.org>.
[ https://issues.apache.org/jira/browse/SANDESHA2C-27?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=12586157#action_12586157 ]
Damitha Kumarage commented on SANDESHA2C-27:
--------------------------------------------
To descrease memory leaks the most important things to consider are
1. Whether module introduced operation contexts and message contexts are freed properly.
2. If Sandesha2/C call axis2_op_ctx_in_use(), axis2_msg_ctx_set_paused(), axis2_msg_ctx_set_keep_alive(),
axis2_op_ctx_increment_ref(), axiom_soap_envelope_increment_ref() to prevent memory freeing of op_ctx and msg_ctx structures at the usual time by the axis2 engine, then we make to
sure if these are actually needed and if they are really needed we need to make sure that those constructs are freed properly when they are used. When I look at the code I can see that above functions are used inappropriately which prevent proper memory freeing. Should investigate on this.
> Do memory leak fixes in Sandesha2c
> ----------------------------------
>
> Key: SANDESHA2C-27
> URL: https://issues.apache.org/jira/browse/SANDESHA2C-27
> Project: Sandesha2/C
> Issue Type: Bug
> Reporter: Damitha Kumarage
>
> Do memory leak fixes in Sandesha2
--
This message is automatically generated by JIRA.
-
You can reply to this email to add a comment to the issue online.
---------------------------------------------------------------------
To unsubscribe, e-mail: sandesha-dev-unsubscribe@ws.apache.org
For additional commands, e-mail: sandesha-dev-help@ws.apache.org