Am Fri, 29 Jul 2016 02:04:42 +0100 schrieb Wols Lists <antlists@xxxxxxxxxxxxxxx>: > On 29/07/16 01:50, Eric Wheeler wrote: > > Hello all, > > > > With the many SSD caching layers being developed (bcache, dm-cache, > > dm-writeboost, etc), how could we flag a bio from userspace to > > indicate whether the bio is preferred to hit spinning disks instead > > of an SSD? > > > > Unnecessary promotions, evections, and writeback increase the write > > burden on the caching layer and burns out SSDs too fast (TBW), thus > > requring equipment replacement. > > What's the spec of these devices? How long are they expected to last? > > Other recent posts on this (linux-raid) mailing list refer to tests on > SSDs that indicates their typical life is way beyond their nominal > life, and that in normal usage they are actually likely to outlive > "spinning rust". Well, using SSD as a caching layer is probably everything else but normal usage. Caching involves writing a lot of data given the fact that it usually backs huge storage pools and is there to eliminate/hide inefficient usage patterns of rotational media. Caching layers like bcache do its best to turn writes to the device into optimal write patterns at best - but still, it writes a lot of data. > http://techreport.com/review/24841/introducing-the-ssd-endurance-experiment > > http://techreport.com/review/27909/the-ssd-endurance-experiment-theyre-all-dead/3 > > Looking at the results, the FIRST drives only started failing once > they'd written some 700 Terabytes. How long is it going to take you to > write that much data over a SATA3 link? I had a Crucial MX100 128GB SSD as bcache (writeback mode) in the past. It lasted exactly 12-13 months when I was forced to replace it: The lifetime of the SSD reached around 95% according to smartctl. The specs say, it has a lifetime guarantee of 85TB written. I guess it could have lasted longer due to bcache's optimized access patterns - but I didn't want to test it. Bigger SSDs usually last a lot longer, so now I'm using a Samsung 850 Evo 500GB which has a guarantee of 150TBW. According to tests, it probably goes a lot higher before failing (like 250TBW estimated) but I don't want to test this is writeback mode. So the idea of reducing writes to the caching layer, or better eliminate useless writes altogether, would be very welcome. -- Regards, Kai Replies to list-only preferred. -- To unsubscribe from this list: send the line "unsubscribe linux-bcache" in the body of a message to majordomo@xxxxxxxxxxxxxxx More majordomo info at http://vger.kernel.org/majordomo-info.html
