RE: New failures for pcp-2.3.0-11

["Zhang, Sonic" <sonic.zhang@xxxxxxxxx>]

Hi,

        Yes, it the problem of rsh.

        Thanks.

        Sonic Zhang


-----Original Message-----
From: Keith Owens [mailto:kaos@xxxxxxx]
Sent: 2003?2?17? 12:04
To: Zhang, Sonic; PCP (E-mail)
Subject: Re: New failures for pcp-2.3.0-11 


On Mon, 17 Feb 2003 14:31:06 +1100, 
Ken McDonell <kenmcd@xxxxxxxxxxxxxxxxx> wrote:
>Never seen this before.  Since we do not explicitly use poll() in the PCP
>source, my guess would be that this is some sort of local networking
>problem associated with rsh or rcp between the QA machines.
>
>On Mon, 17 Feb 2003, Zhang, Sonic wrote:
>>      poll: protocol failure in circuit setup

The message comes from libc::rcmd_af (man rcmd).  That function uses
poll() to check if rsh to the target machine is working.  AFAICT the
message is issued when the target command dies prematurely.