[581] in SIPB-AFS-requests

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

rosebud: restarted 12/22 @ 13:23

daemon@ATHENA.MIT.EDU (probe@MIT.EDU)
Sun Dec 22 13:28:14 1991

From: probe@MIT.EDU
Date: Sun, 22 Dec 91 13:27:41 -0500
To: sipb-afsreq@MIT.EDU

I restarted rosebud's service at this time, because all that was running
was the fileserver and volserver.  Apparently, during the 4am restart,
the bosserver never restarted, nor did any of its children processes.

I will investigate further, but I suspect the easiest way to avoid the
problem is not to have the weekly restart enabled (the rel-eng cell has
been running in thise mode for months without a problem).  I suspect that
the rios is so fast that when the bosserver is restarting that the port
it was listening on is still in time_wait and therefore the next bosserver
can't bind to the port.

-Richard

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