Hey Joe,
We have tried a few different things wrt the C++ clients and thrift. Just
putting out some of out thoughts here.
First, we used the existing Thrift proxy as a separate tier (Thrift proxy
tier). The issue there was that we just didn't get enough throughput (for
various reasons). Indepeden
My 2 cents - whatever branch we decide to put out as 1.0, I think we
should have a stability/testing phase without adding too many features, so
that it is pretty stable to end users.
- Karthik
On 11/16/11 3:57 PM, "Andrew Purtell" wrote:
>> From: Stack
>
>> To: d...@hbase.apache.org; Andrew
Mark,
If the NN goes down the entire HBase cluster will be halted. If the
HMaster is down, you will not be able to get regions assigned (in case any
regionserver goes down).
So the best thing to do is start NN and Hmaster on another node, do you
upgrade and move them back.
Thanks
Karthik
On 11
@Mark - like Doug said its dependent on the hardware SKU and load type. We
(at FB) are using 300 handlers on a box with 16 cores and 48G (24G for RS).
On 11/15/11 9:43 AM, "Mark" wrote:
>Perhaps section 2.8.2.3 of the hbase book should be updated then?
>
>On 11/15/11 9:17 AM, Doug Meil wrote:
>>
<< ...mod the hash with the number of machines I have... >>
This means that the data will change with the number of machines - so all
your data will map to different regions if you add a new machine to your
cluster.
<< What I do not understand is the advantages/disasvantages of having
regions t
Hey Matthew,
The only way to increase the number of reducers is to have more regions -
each reducer produces an output per region, so the number of reducers ==
number of regions.
Thanks
Karthik
On 10/18/11 2:00 AM, "Matthew Tovbin" wrote:
>Hello, Guys,
>
>I'm willing to bulk load data from hd