If you profiled it, where did you find the time being spent? St.Ack
On Tue, Mar 5, 2013 at 11:00 PM, kiran <kiran.sarvabho...@gmail.com> wrote: > Yes we profiled the gets after completing the scan only. We have 12 > regionservers. Each table i am doing gets has 40 regions. > > > On Wed, Mar 6, 2013 at 12:21 PM, Stack <st...@duboce.net> wrote: > > > On Tue, Mar 5, 2013 at 8:36 PM, kiran <kiran.sarvabho...@gmail.com> > wrote: > > > > > Dear All, > > > > > > I had some miserable experience with gets (batch gets) in hbase. I have > > two > > > tables with different rowkeys, columns are distributed across the two > > > tables. > > > > > > Currently what I am doing is scan over one table and get all the > rowkeys > > in > > > the first table matching my filter. Then issue a batch get on another > > table > > > to retrieve some columns. But even for 20 gets, the performance is like > > > miserable (almost a second or two for 20 gets which is not acceptable). > > > But, scanning even on few thousands of rows is getting completed in > > > milliseconds. > > > > > > > > What happens if you do those 20 batch gets in isolation, apart from the > > scan? Do you still have miserable performance? > > > > What is your cluster setup like? How many nodes? How many regions? > > > > > > > > > My concern is for about 20 gets if it takes second or two, > > > How can it scale ?? > > > > > > > Normally it does not take this long. Lets figure out what is going on on > > your setup. > > > > St.Ack > > > > > > -- > Thank you > Kiran Sarvabhotla > > -----Even a correct decision is wrong when it is taken late >