On Thu, Oct 18, 2018 at 12:31 PM Matt Benjamin <mbenjami@xxxxxxxxxx> wrote: > > spotted a couple of things > > 1. "The object's manifest has a tail": could you explain this? > Ah, I think the sentence there was cut and some information that I was planning to provide got lost. I think I wanted to say that the objects' manifest has a tail field that points to the tail's placement target. > 2. >We should probably make it so that when head and tail are being placed > > on different placement targets, the head will not contain any data, > > other than the object’s metadata. > > I'd prefer to have the option to retain the small-object optimization, > Casey noted this too (conversation); why is this ruled out? > Of course, and it's not ruled out. Just need to keep in mind that the current head size of 4MB is in some use cases is bigger than the size of most of the objects in the system, so many users will not want to have all that data duplicated. It's a tradeoff. We can maybe make it easier to configure, so that users make their own decision there. Yehuda > regards, > > Matt > > -- > > Matt Benjamin > Red Hat, Inc. > 315 West Huron Street, Suite 140A > Ann Arbor, Michigan 48103 > > http://www.redhat.com/en/technologies/storage > > tel. 734-821-5101 > fax. 734-769-8938 > cel. 734-216-5309
