[rds-devel] FW: RDS -- RDMA out of order

[Tang, Changqing]

Since it is in  loop of iteration, some iteration has no failure, other iteration has processes randomly killed and replaced with new processes,  it is hard to which dmesg
Message belongs to which iteration.

But dmesg has either:

RDS/IB: send completion on 172.31.64.31 had status 10, disconnecting and reconnecting
RDS/IB: connected to 172.31.64.31 version 3.1

Or:

RDS/IB: connected to 172.31.64.31 version 3.1
RDS/IB: connected to 172.31.64.31 version 3.1
RDS/IB: connected to 172.31.64.31 version 3.1

--CQ


-----Original Message-----
From: Andy Grover [mailto:andy.grover at oracle.com]
Sent: Wednesday, May 05, 2010 8:44 PM
To: Tang, Changqing
Cc: RDS Devel
Subject: Re: [rds-devel] FW: RDS -- RDMA out of order

On 05/05/2010 05:38 AM, Tang, Changqing wrote:
> Andy, Normally I see completion notification with remote error, then
> a lot of dropped notifications. But sometimes during normal run, I
> see a couple dropped notifications only. Do You have explanation why
> there is only dropped notification without a leading error
> notification ?

What messages in dmesg are you seeing along with both cases?

Regards -- Andy