Hi Nick, Our latency probe results (4kB rados bench) didn't change noticeably after converting a test cluster from FileStore (sata SSD journal) to BlueStore (sata SSD db). Those 4kB writes take 3-4ms on average from a random VM in our data centre. (So bluestore DB seems equivalent to FileStore journal for small writes). Otherwise, our other monitoring (osd log analysis) shows that the vast majority of writes are under 32kB, and the average write size is 42kB (with a long tail out to 4MB). So... do you think is this a *real* issue that would impact user observed latency, given that they are mostly doing small writes? (maybe your environment is very different?) I'm not saying that tuning the deferred write threshold up wouldn't help, but it's not obvious that deferring writes is better on the whole. I'm curious what was the original rationale for 32kB? Cheers, Dan On Tue, May 1, 2018 at 10:50 PM, Nick Fisk <nick@xxxxxxxxxx> wrote: > Hi all, > > > > Slowly getting round to migrating clusters to Bluestore but I am interested > in how people are handling the potential change in write latency coming from > Filestore? Or maybe nobody is really seeing much difference? > > > > As we all know, in Bluestore, writes are not double written and in most > cases go straight to disk. Whilst this is awesome for people with pure SSD > or pure HDD clusters as the amount of overhead is drastically reduced, for > people with HDD+SSD journals in Filestore land, the double write had the > side effect of acting like a battery backed cache, accelerating writes when > not under saturation. > > > > In some brief testing I am seeing Filestore OSD’s with NVME journal show an > average apply latency of around 1-2ms whereas some new Bluestore OSD’s in > the same cluster are showing 20-40ms. I am fairly certain this is due to > writes exhibiting the latency of the underlying 7.2k disk. Note, cluster is > very lightly loaded, this is not anything being driven into saturation. > > > > I know there is a deferred write tuning knob which adjusts the cutover for > when an object is double written, but at the default of 32kb, I suspect a > lot of IO’s even in the 1MB area are still drastically slower going straight > to disk than if double written to NVME 1st. Has anybody else done any > investigation in this area? Is there any long turn harm at running a cluster > deferring writes up to 1MB+ in size to mimic the Filestore double write > approach? > > > > I also suspect after looking through github that deferred writes only happen > when overwriting an existing object or blob (not sure which case applies), > so new allocations are still written straight to disk. Can anyone confirm? > > > > PS. If your spinning disks are connected via a RAID controller with BBWC > then you are not affected by this. > > > > Thanks, > > Nick > > > _______________________________________________ > ceph-users mailing list > ceph-users@xxxxxxxxxxxxxx > http://lists.ceph.com/listinfo.cgi/ceph-users-ceph.com > _______________________________________________ ceph-users mailing list ceph-users@xxxxxxxxxxxxxx http://lists.ceph.com/listinfo.cgi/ceph-users-ceph.com
 |