From mboxrd@z Thu Jan 1 00:00:00 1970 From: NeilBrown Subject: Re: [PATCH V4 00/13] MD: a caching layer for raid5/6 Date: Mon, 20 Jul 2015 10:03:45 +1000 Message-ID: <20150720100345.63302007@noble> References: <20150710040847.GA1408097@devbig257.prn2.facebook.com> <20150710143656.4ee7e647@noble> <20150715004455.GA1024035@devbig257.prn2.facebook.com> <20150715121234.22803890@noble> <20150715031615.GA1292235@devbig257.prn2.facebook.com> <20150715140641.3c61cf78@noble> <20150715194927.GA3502691@devbig257.prn2.facebook.com> <20150716091653.7b970b32@noble> <20150716000739.GA4041321@devbig257.prn2.facebook.com> <20150716112217.3e2e685f@noble> <20150716041305.GA283306@devbig257.prn2.facebook.com> <20150716160711.08c37fc1@noble> <21927.51382.8600.700720@quad.stoffel.home> Mime-Version: 1.0 Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 7bit Return-path: In-Reply-To: <21927.51382.8600.700720@quad.stoffel.home> Sender: linux-raid-owner@vger.kernel.org To: John Stoffel Cc: Shaohua Li , linux-raid@vger.kernel.org, songliubraving@fb.com, hch@infradead.org, dan.j.williams@intel.com, Kernel-team@fb.com List-Id: linux-raid.ids On Thu, 16 Jul 2015 11:07:34 -0400 "John Stoffel" wrote: > > So.... how will wrapping of sequence numbers be handled in the cache > metadata on disk? They can't grow for ever... Can't they? A 64bit number incremented every millisecond can grow for half a billion years. As far as I'm concerned, that is forever. But even if we used 32bit numbers, it is very easy to handle looping providing you have few than half of the numbers in use at any time. So as long as we don't have more than 2 billion active transactions in the log at once, there is no problem. The kernel does some internal time keeping using "jiffies" which, on 32bit systems, cycle about every fortnight. A small amount of care is enough that this doesn't cause problem. NeilBrown