You are viewing a plain text version of this content. The canonical link for it is here.
Posted to dev@tomcat.apache.org by Mark Thomas <ma...@apache.org> on 2006/11/25 17:39:40 UTC
StandardManager behaviour when webapp stopped
Bug 40593 [1] has raised the following question:
When a web app with valid sessions is stopped and the StandardManager
is configured not to serialize the sessions what, if anything, should
the container do in this case?
As far as I can tell from the spec, the container behaviour in these
circumstances is undefined. Is there a good reason not to invalidate
the active sessions in this case?
Thoughts?
Mark
[1] http://issues.apache.org/bugzilla/show_bug.cgi?id=40593
---------------------------------------------------------------------
To unsubscribe, e-mail: dev-unsubscribe@tomcat.apache.org
For additional commands, e-mail: dev-help@tomcat.apache.org
Re: StandardManager behaviour when webapp stopped
Posted by Mark Thomas <ma...@apache.org>.
Remy/Filip,
Thanks for your input. I'll update the bug report with a brief summary
and come back to this later.
Mark
Filip Hanik - Dev Lists wrote:
> You can still cluster just using AOP without ever involving the manager,
> hence such solutions would be affected by this new behavior.
> at most, when a webapp is shutdown, you can call passivate, but not
> invalidate the sessions.
> Filip
>
> Remy Maucherat wrote:
>> Filip Hanik - Dev Lists wrote:
>>> it will cause problems for clustered stuff, not sure that container
>>> shutdown means session invalidation.
>>> for example, if you invalidate the session, such a change would
>>> trickle through the cluster, and that might not be intentional.
>>> I suggest we leave the behavior as is, or at least configurable if we
>>> mean to change it
>>
>> The clustered manager most likely would have the opportunity to
>> override the code which would do the expiration, I think.
>>
>> Rémy
>>
>> ---------------------------------------------------------------------
>> To unsubscribe, e-mail: dev-unsubscribe@tomcat.apache.org
>> For additional commands, e-mail: dev-help@tomcat.apache.org
>>
>>
>>
>
>
>
> ---------------------------------------------------------------------
> To unsubscribe, e-mail: dev-unsubscribe@tomcat.apache.org
> For additional commands, e-mail: dev-help@tomcat.apache.org
>
>
---------------------------------------------------------------------
To unsubscribe, e-mail: dev-unsubscribe@tomcat.apache.org
For additional commands, e-mail: dev-help@tomcat.apache.org
Re: StandardManager behaviour when webapp stopped
Posted by Filip Hanik - Dev Lists <de...@hanik.com>.
You can still cluster just using AOP without ever involving the manager,
hence such solutions would be affected by this new behavior.
at most, when a webapp is shutdown, you can call passivate, but not
invalidate the sessions.
Filip
Remy Maucherat wrote:
> Filip Hanik - Dev Lists wrote:
>> it will cause problems for clustered stuff, not sure that container
>> shutdown means session invalidation.
>> for example, if you invalidate the session, such a change would
>> trickle through the cluster, and that might not be intentional.
>> I suggest we leave the behavior as is, or at least configurable if we
>> mean to change it
>
> The clustered manager most likely would have the opportunity to
> override the code which would do the expiration, I think.
>
> Rémy
>
> ---------------------------------------------------------------------
> To unsubscribe, e-mail: dev-unsubscribe@tomcat.apache.org
> For additional commands, e-mail: dev-help@tomcat.apache.org
>
>
>
---------------------------------------------------------------------
To unsubscribe, e-mail: dev-unsubscribe@tomcat.apache.org
For additional commands, e-mail: dev-help@tomcat.apache.org
Re: StandardManager behaviour when webapp stopped
Posted by Remy Maucherat <re...@apache.org>.
Filip Hanik - Dev Lists wrote:
> it will cause problems for clustered stuff, not sure that container
> shutdown means session invalidation.
> for example, if you invalidate the session, such a change would trickle
> through the cluster, and that might not be intentional.
> I suggest we leave the behavior as is, or at least configurable if we
> mean to change it
The clustered manager most likely would have the opportunity to override
the code which would do the expiration, I think.
Rémy
---------------------------------------------------------------------
To unsubscribe, e-mail: dev-unsubscribe@tomcat.apache.org
For additional commands, e-mail: dev-help@tomcat.apache.org
Re: StandardManager behaviour when webapp stopped
Posted by Filip Hanik - Dev Lists <de...@hanik.com>.
it will cause problems for clustered stuff, not sure that container
shutdown means session invalidation.
for example, if you invalidate the session, such a change would trickle
through the cluster, and that might not be intentional.
I suggest we leave the behavior as is, or at least configurable if we
mean to change it
Filip
Remy Maucherat wrote:
> Mark Thomas wrote:
>> Bug 40593 [1] has raised the following question:
>>
>> When a web app with valid sessions is stopped and the StandardManager
>> is configured not to serialize the sessions what, if anything, should
>> the container do in this case?
>>
>> As far as I can tell from the spec, the container behaviour in these
>> circumstances is undefined. Is there a good reason not to invalidate
>> the active sessions in this case?
>>
>> Thoughts?
>
> I think it would be ok.
>
> Rémy
>
> ---------------------------------------------------------------------
> To unsubscribe, e-mail: dev-unsubscribe@tomcat.apache.org
> For additional commands, e-mail: dev-help@tomcat.apache.org
>
>
>
---------------------------------------------------------------------
To unsubscribe, e-mail: dev-unsubscribe@tomcat.apache.org
For additional commands, e-mail: dev-help@tomcat.apache.org
Re: StandardManager behaviour when webapp stopped
Posted by Remy Maucherat <re...@apache.org>.
Mark Thomas wrote:
> Bug 40593 [1] has raised the following question:
>
> When a web app with valid sessions is stopped and the StandardManager
> is configured not to serialize the sessions what, if anything, should
> the container do in this case?
>
> As far as I can tell from the spec, the container behaviour in these
> circumstances is undefined. Is there a good reason not to invalidate
> the active sessions in this case?
>
> Thoughts?
I think it would be ok.
Rémy
---------------------------------------------------------------------
To unsubscribe, e-mail: dev-unsubscribe@tomcat.apache.org
For additional commands, e-mail: dev-help@tomcat.apache.org