From mboxrd@z Thu Jan 1 00:00:00 1970 References: From: Nemo Content-Type: text/plain; charset=us-ascii In-Reply-To: Message-Id: Date: Thu, 22 Mar 2012 20:54:36 +0100 To: Fans of the OS Plan 9 from Bell Labs <9fans@9fans.net> Content-Transfer-Encoding: 7bit Mime-Version: 1.0 (1.0) Subject: Re: [9fans] fossil archive corruption - solved Topicbox-Message-UUID: 6e77657a-ead7-11e9-9d60-3106f5b1d025 impressive. hats off. -- iphone kbd. excuse typos :) On Mar 22, 2012, at 8:24 PM, Richard Miller <9fans@hamnavoe.com> wrote: >> I've not been able to easily reproduce this problem. >> How did you proceed? > > Not very reproducible, but I found it could be encouraged to happen > (on a sacrificial secondary fossil+venti) with this: > tar c /n/xfossil/sys/src >/dev/null & > while (sleep 60) echo snap -a >>/srv/xfscons > > An easily recognisable consequence when it fails is a BtDir block with > invalid contents. So I peppered the source with diagnostics checking > for that, until I was able to spot exactly where it happened. >