Re: MapReduce performance problem

I didn't want you to think that you've been forgotten, but I've been swamped getting ready to head out of the country for 2 weeks on a company trip. You're in good hands with the list, though. --- Jeremiah Peschka - Founder, Brent Ozar Unlimited MCITP: SQL Server 2008, MVP Cloudera Certified Devel

Re: Tuning a Riak cluster.

Not forgotten you Kevin, out of the country at the moment, will get back to you soon as I can Regards Richard On Tue, Feb 26, 2013 at 4:06 PM, Kevin Burton wrote: > Here is the config that I am using: > > ** ** > > {driver, > > basho_bench_driver_2i}. > > ** ** > > {operations,

Riak 1.3 wont start on Macos

Hi Riak Users, Here is what I have: ::riak]ulimit -a core file size (blocks, -c) 0 data seg size (kbytes, -d) unlimited file size (blocks, -f) unlimited max locked memory (kbytes, -l) unlimited max memory size (kbytes, -m) unlimited open files

Re: Understanding read_repairs

I just tried and it worked for me, here's what I see: ➜ riak_kv git:(master) ./rebar --version rebar 2.0.0 R14B03 20120816_132849 git 1c98f6c ➜ riak_kv git:(master) erl Erlang R15B01 (erts-5.9.1) [source] [64-bit] [smp:8:8] [async-threads:0] [kernel-poll:false] I also tried erlang 14b04 and it

Re: MapReduce scalability

Thanks Christian and BR Rune for your answers, we'll reconsider how we did our processing since obviously we badly react to tombstones when our MR query meet them. Best Regards, Bernard Le 28/02/2013 16:17, Christian Dahlqvist a écrit : Hi Bernard, The description in the documentation is e

Re: Wiping data directory break vnodes with read_block_error

Hello Jonas, I think the issue is that you actually wiped out too much by removing the entire /var/lib/riak directory. Typically when resetting the data directories on a Riak node, you'd remove the contents of only those backend directories which correspond to the ones defined in your app.config.

Re: MapReduce scalability

On Thu, Feb 28, 2013 at 5:53 AM, wrote: > The recommended way to implement efficient multiget is to perform normal > GET operations in parallel. If you are retrieving 20 objects, you don't > necessarily need to do all 20 GETs in parallel, but could set it up to use > perhaps 3 or 4 connections. I

Re: MapReduce scalability

A general comment on Riak's Reduce scalability. Map scales fine but Reduce doesn't. The single coordinating node does the final reducing, so latency from reduces will not improve from adding more nodes. MR jobs where lots of data is sent from map to reduce, will also get the coordinating node

{error, notfound} items in mapreduce during ownership handoff

Yesterday I added a new node to my cluster. During the time when ownership handoff was happening (several hours of work) mapreduce map functions were receiving {error, notfound} as inputs. My Erlang mapred functions weren't designed to handle this. They hadn't encountered this before during normal

Re: Understanding read_repairs

Jared, running make in riak_kv dies [0] for me :-/ Best Sebastian [0]: https://gist.github.com/tisba/5057828 On 22.02.2013, at 19:56, Jared Morrow wrote: > compile riak_kv with just 'make' ___ riak-users mailing list riak-users@lists.basho.com h

Re: Lost keys

Hey Pavel, Sorry to hear you're running into trouble. Here's a couple of things that will help us help you out: Can you provide some more info, or code used to load Riak and Redis? What version of the Python client are you using? Have you tried retrieving the missing keys with another client,

Re: MapReduce scalability

Hi Bernard, The description in the documentation is entirely accurate and not at all purely theoretical. Riak will automatically select a covering set of vnodes/partitions that hold the data set required to complete the job. All physical nodes may therefore net need to participate in the job. W

Re: MapReduce scalability

Hi Christian, At http://docs.basho.com/riak/1.3.0/references/appendices/MapReduce-Implementation/ , one can read "...any Riak node can also coordinate a MapReduce query by sending a map-step evaluation request directly to the node responsible for maintaining the input data. Map-step results a

Re: MapReduce scalability

Hi Boris, Apart from not scaling quite as well as straight K/V access, emulating multiGET through MapReduce also has another significant drawback. MapReduce has no concept of quorum reads, and only work on a single copy of the data, which can be thought of basically as a read with R=1 that does

Re: Wiping data directory break vnodes with read_block_error

I realize now that read_block_error, <<"0">> means that there was not an error... I still get problems when running my integration tests: {insufficient_vnodes,0,need,2} status code 0 Could it be a timing issue where I start my integration tests too early after 'riak start'? On 28 February 2013