[1105] in SIPB-AFS-requests

home	help	back	first	fref	pref	prev	next	nref	lref	last	post

Possible AFS problem

daemon@ATHENA.MIT.EDU (Matt Braun)
Tue Aug 10 12:43:39 1993

To: sipb-afsreq@mit.edu, zone-cell@mit.edu, probe@mit.edu
Date: Tue, 10 Aug 93 12:43:16 EDT
From: Matt Braun <mhbraun@mit.edu>


Apparently bos listhosts isn't always truthful about what the current
primaries are in the cell.


I did the following: 

	seiko# cat CellServDB
	>zone.mit.edu   #zone cell
	18.177.0.20     #casio.mit.edu
	18.177.0.21     #timex.mit.edu
	18.177.0.22     #seiko.mit.edu

	seiko# ed CellServDB
	120
	/seiko/p
	18.177.0.22     #seiko.mit.edu
	d
	w
	89
	q
	
	seiko# diff CellServDB CellServDB.orig
	3a4
	> 18.177.0.22     #seiko.mit.edu
	
	seiko# bos listhosts seiko -c zone
	bos: no such entry (getting tickets)
	bos: running unauthenticated
	Cell name is zone.mit.edu
	    Host 1 is casio.mit.edu
	    Host 2 is timex.mit.edu
	    Host 3 is seiko.mit.edu
	seiko#
	seiko# bos shutdown seiko -c zone -wait -local
	seiko# bos restart seiko -all -c zone  -local
	seiko#
	seiko# bos status seiko
	bos: no such entry (getting tickets)
	bos: running unauthenticated
	Instance fs, currently running normally.
	    Auxiliary status is: file server running.
	Instance vlserver, currently running normally.
	Instance ptserver, currently running normally.
	
	seiko# bos listhosts seiko -c zone
	bos: no such entry (getting tickets)
	bos: running unauthenticated
	Cell name is zone.mit.edu
	    Host 1 is casio.mit.edu
	    Host 2 is timex.mit.edu
	    Host 3 is seiko.mit.edu
	seiko# cat CellServDB
	>zone.mit.edu   #zone cell
	18.177.0.20     #casio.mit.edu
	18.177.0.21     #timex.mit.edu
	
Now I followed this with:
	
	seiko# bos restart seiko -all -bosserver  -c zone -local
	seiko# bos listhosts seiko -c zone
	bos: no such entry (getting tickets)
	bos: running unauthenticated
	Cell name is zone.mit.edu
	    Host 1 is casio.mit.edu
	    Host 2 is timex.mit.edu
	
which no longer is lying..

	
Now apparently the bos server is caching the infomration about the CSDB even
across a bos restart...

Is this a bug ? or should we never be ed'ing the CSDB by hand ? and just use
bos removehost and addhost for such matters (Richard: no need to flame about
changing the CSDB here) 

				Matt
	
	
	

home	help	back	first	fref	pref	prev	next	nref	lref	last	post