pcp
[Top] [All Lists]

Re: Bug in case 232 is root caused.

To: Keith Owens <kaos@xxxxxxx>
Subject: Re: Bug in case 232 is root caused.
From: Max Matveev <makc@xxxxxxx>
Date: Sat, 25 Jan 2003 11:29:48 +1100
Cc: kenmcd@xxxxxxxxxxxxxxxxx, "Zhang, Sonic" <sonic.zhang@xxxxxxxxx>, "PCP (E-mail)" <pcp@xxxxxxxxxxx>
In-reply-to: <13501.1043452157@xxxxxxxxxxxxxxxxxxxxx>
References: <Pine.LNX.4.44.0301250919450.25555-100000@xxxxxxxxxxxxxxxxxxxxxxxxx> <13501.1043452157@xxxxxxxxxxxxxxxxxxxxx>
Sender: pcp-bounce@xxxxxxxxxxx
>>>>> "kaos" == Keith Owens <kaos@xxxxxxx> writes:

 kaos> On Sat, 25 Jan 2003 09:21:40 +1100 (EST), 
 kaos> kenmcd@xxxxxxxxxxxxxxxxx wrote:
 >> On Fri, 24 Jan 2003, Zhang, Sonic wrote:
 >>> I find the root cause of the bug in case 232.
 >>> NFS service daemon should be running before test this case.
 >> 
 >> Excellent.  Is there some test we could make for this, so that
 >> the qa script could issue a more useful error message ... the
 >> 232.out.bad file was not that helpful in this case!

 kaos> Quick and dirty, 99% reliable.

 kaos> [ ! -s /var/lib/nfs/etab -a ! -s /etc/xtab ] && \
 kaos>  echo Either nfs is not running or nothing is exported

 kaos> That will only fail if nfs has been killed -9 and not restarted, when
 kaos> it will give a false positive.

 kaos> To make it 100% reliable would take a program which contacted nfsd and
 kaos> asked for a list of exported filesystems.  Messy, needs portmapper code
 kaos> as well.
FYI, list of exported FSes comes from mountd, not from nfs server. To
test if server it up or not, rpcinfo -u foo 100003 is a better option.

max

<Prev in Thread] Current Thread [Next in Thread>