[ofa-general] RDS flow control

[Richard Frank]

 From Oracle's perspective I think you can punt on the cluster rolling 
upgrade aspect of this.

It's likely that Oracle will be running a newer version of RDS with flow 
control - or all nodes will be on an older version.

As long as different versions of drivers interact without crashing the 
node - and report a mismatch in protocol - we're probably OK.

Assuming the flow control turns out to not impact performance - then we 
should just remove the old RNR code.

If we're ready = I give this version of the driver to our performance 
folks - hopefully they can give a bash in the next week or so..


Olaf Kirch wrote:
> On Tuesday 20 May 2008 22:45:22 Jon Mason wrote:
>   
>> Works well on my setup.
>>     
>
> Good to hear!
>
>   
>> With proper flow control, there should no longer be a need for rnr_retry (as there
>> should always be a posted recv buffer waiting for the incoming data).  I did a
>> quick test and removed it and everything seemed to be happy on my rds-stress run.
>>     
>
> I would like to make the setting of the RNR retry/timeout conditional on
> whether both ends of the connection support flow control or not - we need
> to think of rolling upgrades of a cluster, so mixed environments just have
> to work. Unfortunately, the RNR retry count is set prior to establishing
> the connection, before we even know whether the remote is capable of doing
> flow control.
>
> Is there a way of changing the RNR retry count back to 0 after establishing
> the connection?
>
> Olaf
>