mina-dev mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From "Niklas Gustavsson (JIRA)" <j...@apache.org>
Subject [jira] Commented: (FTPSERVER-304) The Idle Timeout set in the listener configuration does not have any effect
Date Mon, 25 May 2009 19:18:45 GMT

    [ https://issues.apache.org/jira/browse/FTPSERVER-304?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=12712776#action_12712776

Niklas Gustavsson commented on FTPSERVER-304:

> Well...The impression I get by looking at the configuration options/reading
> the documentation is that the "idle-timeout" parameter is *NOT* the
> max-idle-time, but is the *default* timeout.
> Below is the snippet from the documentation:
> idle-timeout - The number of seconds before an inactive client is
> disconnected

The idle-timeout has always (as far as I know) been the maximum idle time that is allowed
and I even think the documentation says so (although the documentation does not cover the
inheritance scheme which we need to add as I referred to above).

> Basically we should be able to support the following:
> * There should be a way on "User" records which tells us that the user gets
> the default timeout which is defined on the listener. Since the timeout on
> User records is an integer, it has to be a value that is less than 0. 0
> would mean infinite and a positive number implies a finite timeout.

The current design supports this (by setting the user limit to 0).

> * Ability to override the timeout on a per user basis. This could be less
> than the default timeout, greater than the default timeout or infinite
> timeout. For example, all "admin" level users might want to have infinite
> timeout even though the default timeout is set to 300.

This we do not currently support, and personally I don't think we should. I think the limit
on the listener should always be the upper limit. If you need to differentiate your users,
do it below the listener limit.

> I'm not sure if it is of much advantage to support 'max-idle-timeout' (if we
> do the above two), but if you think it is, then we should perhaps add
> another configuration option for max-idle-timeout keeping the existing one
> as default timeout.

I think the current design is complex as it is, I don't think we should complicate it any
further. But that's just my opinion. I'm known for changing my mind if people speak up :-)

> The Idle Timeout set in the listener configuration does not have any effect
> ---------------------------------------------------------------------------
>                 Key: FTPSERVER-304
>                 URL: https://issues.apache.org/jira/browse/FTPSERVER-304
>             Project: FtpServer
>          Issue Type: Bug
>          Components: Core
>    Affects Versions: 1.0.1
>            Reporter: Sai Pullabhotla
>             Fix For: 1.0.2
> As reported on the mailing list by Johannes, I confirmed that the idle timeout set on
the listener factory did not have any effect. The connection/session is still good long after
the specified idle timeout. As far as I can remember this used to work fine in pre-1.0 releases.
We also have to make sure that the idle timeout on the data connections works. 

This message is automatically generated by JIRA.
You can reply to this email to add a comment to the issue online.

View raw message