[MINA 3.0] Some thoughts on Debug tool

[Ashish <pa...@gmail.com>]

Guys,

I just added some thoughts on the Debug tool on MINA 3.0 design page
(https://cwiki.apache.org/confluence/display/MINA/MINA+3.0+design).
Ideas are better on wiki than in my mind.

JMX would be the best way to get the data from running App, as we
shall be able to extract more data from running JVM.

On the initial thoughts, we can capture Statistical and Configuration
information in the tool, which shall help see the system in action.

Few items indentified are

Configuration - (Probably would be good if we can allow change in
config from the same UI as well as expose JMX API's)
Chain Config
Acceptor
Selector
Strategy ??

Stats
Session's
Session data
Heaps Usage
???

I believe our users can provide more info on what might help them
better manage/debug their systems.

Thoughts/Comments/Suggestion?

thanks
ashish

Re: [MINA 3.0] Some thoughts on Debug tool

[Ashish <pa...@gmail.com>]

>>
>> On the initial thoughts, we can capture Statistical and Configuration
>> information in the tool, which shall help see the system in action.
>>
>> Few items indentified are
>>
>> Configuration - (Probably would be good if we can allow change in
>> config from the same UI as well as expose JMX API's)
>> Chain Config
>> Acceptor
>> Selector
>> Strategy ??
>
> I would add Log level. It would be cool if we can switch it dynamically.

This would be a great feature
+1


>>
>> Stats
>> Session's
>> Session data
>> Heaps Usage
>> ???
>
> It would be very useful to gather information related to the queues we are
> using in many places.
>
> Also inactive sessions, sessions duration, etc

+1, all the session related data should be visible after selecting the
session id, including data transfer, the chain composition, Queue size
etc

>>
>> I believe our users can provide more info on what might help them
>> better manage/debug their systems.
>>
>> Thoughts/Comments/Suggestion?
>
> Everything that can help the user to have a better insight on what's going
> on is a good thing to have.

Will add these to the wiki page so that we don't loose them.

Anyone would like to see specific data that would help them?

I think we can send one mail to users ML as well. wdyt?

Re: [MINA 3.0] Some thoughts on Debug tool

[Emmanuel Lecharny <el...@gmail.com>]

  On 8/19/10 3:41 PM, Ashish wrote:
> Guys,
>
> I just added some thoughts on the Debug tool on MINA 3.0 design page
> (https://cwiki.apache.org/confluence/display/MINA/MINA+3.0+design).
> Ideas are better on wiki than in my mind.
>
> JMX would be the best way to get the data from running App, as we
> shall be able to extract more data from running JVM.
Absolutely !
> On the initial thoughts, we can capture Statistical and Configuration
> information in the tool, which shall help see the system in action.
>
> Few items indentified are
>
> Configuration - (Probably would be good if we can allow change in
> config from the same UI as well as expose JMX API's)
> Chain Config
> Acceptor
> Selector
> Strategy ??

I would add Log level. It would be cool if we can switch it dynamically.
> Stats
> Session's
> Session data
> Heaps Usage
> ???

It would be very useful to gather information related to the queues we 
are using in many places.

Also inactive sessions, sessions duration, etc
> I believe our users can provide more info on what might help them
> better manage/debug their systems.
>
> Thoughts/Comments/Suggestion?
Everything that can help the user to have a better insight on what's 
going on is a good thing to have.


-- 
Regards,
Cordialement,
Emmanuel Lécharny
www.iktek.com

Re: [MINA 3.0] Some thoughts on Debug tool

[Mark Webb <el...@gmail.com>]

I am all for using JMX.  I have used JMX in a few MINA-based
applications to gather performance numbers and worked out really well.


On Thu, Aug 19, 2010 at 9:41 AM, Ashish <pa...@gmail.com> wrote:
> Guys,
>
> I just added some thoughts on the Debug tool on MINA 3.0 design page
> (https://cwiki.apache.org/confluence/display/MINA/MINA+3.0+design).
> Ideas are better on wiki than in my mind.
>
> JMX would be the best way to get the data from running App, as we
> shall be able to extract more data from running JVM.
>
> On the initial thoughts, we can capture Statistical and Configuration
> information in the tool, which shall help see the system in action.
>
> Few items indentified are
>
> Configuration - (Probably would be good if we can allow change in
> config from the same UI as well as expose JMX API's)
> Chain Config
> Acceptor
> Selector
> Strategy ??
>
> Stats
> Session's
> Session data
> Heaps Usage
> ???
>
> I believe our users can provide more info on what might help them
> better manage/debug their systems.
>
> Thoughts/Comments/Suggestion?
>
> thanks
> ashish
>