Queries on the slave could be one reason. However, I see that in the perf test on the wiki also shows the same transfer speed (with rsync too!). Not sure whats up.
2009/10/12 Noble Paul നോബിള് नोब्ळ् <noble.p...@corp.aol.com> > Did you try w/o firing queries on the slave? > > On Sun, Oct 11, 2009 at 6:05 AM, Mark Miller <markrmil...@gmail.com> > wrote: > > > > > > On a drive that can do 40+ that's getting query load might have it's > writes > > knocked down to that? > > > > - Mark > > > > http://www.lucidimagination.com (mobile) > > > > On Oct 10, 2009, at 6:41 PM, Mark Miller <markrmil...@gmail.com> wrote: > > > >> Anyone know why you would see a transfer speed of just 10-20MB over a > >> gigbit network connection? > >> > >> Even with standard drives, I would expect to at least see around 40MB. > >> Has anyone seen over 10-20 using replication? > >> > >> Any ideas on what the bottleneck should be? I think even a standard > >> drive can do writes of a bit of 40MB/s, and certainly reads over that. > >> > >> Thoughts? > >> > >> -- > >> - Mark > >> > >> http://www.lucidimagination.com > >> > >> > >> > > > > > > -- > ----------------------------------------------------- > Noble Paul | Principal Engineer| AOL | http://aol.com > -- Regards, Shalin Shekhar Mangar.
