Re: idea drive layout - 4 drives + RAID question

2012-10-30 Thread Ran User
Is there a concern of a large falloff in commit log write performance (sequential) when sharing 2 drives (RAID 1) with the OS (os and services writing their own logs, etc)? Do you expect the hit to be marginal? On Tue, Oct 30, 2012 at 7:58 PM, aaron morton wrote: > We also have 4-disk nodes, an

Re: idea drive layout - 4 drives + RAID question

2012-10-30 Thread aaron morton
> We also have 4-disk nodes, and we use the following layout: > 2 x OS + Commit in RAID 1 > 2 x Data disk in RAID 0 +1 You are replicating data at the application level and want the fastest possible IO performance per node. > You can already distribute the > individual Cassandra column familie

Re: Wrong data after rolling restart

2012-10-30 Thread Rob Coli
On Mon, May 21, 2012 at 7:08 AM, Alain RODRIGUEZ wrote: > Here are my 2 nodes starting logs, I hop it can help... > > https://gist.github.com/2762493 > https://gist.github.com/2762495 I see in these logs that you replay 2 mutations per node, despite doing nodetool drain before restarting. However

Re: my ubunutu system has "fuseblk" filesystem, would it go faster if I changed it to EXT4?

2012-10-30 Thread Tupshin Harper
Yes. Any fuse filesystem is going to be substantially slower than a massive one like ext4. -Tupshin On Oct 30, 2012 2:09 PM, "Brian Tarbox" wrote: > I got some new ubuntu servers to add to my cluster and found that the file > system is "fuseblk" which really means NTFS. > > All else being equal

Re: Benifits by adding nodes to the cluster

2012-10-30 Thread weiz
One follow up questions. I have been told that it's much easier to scale the cluster by doubling the number of nodes, since no token changed needed on the existing nodes. But if the number of nodes is substantial, it's not realistic to double it every time. How easy is to add let's say 3 additional

my ubunutu system has "fuseblk" filesystem, would it go faster if I changed it to EXT4?

2012-10-30 Thread Brian Tarbox
I got some new ubuntu servers to add to my cluster and found that the file system is "fuseblk" which really means NTFS. All else being equal would I expect to get any performance boost if I converted the file system to EXT4? Edward Capriolo's "Cookbook" book seems to suggest so. Thanks. Brian T

Re: Throughput decreases as latency increases with YCSB

2012-10-30 Thread aaron morton
Out of interest can you quantify the throughput reduction ? It looks like less than 10%. Nice to see it roughly correspond to SEDA throughput :) > Check how many concurrent real requests you have vs size of thread pools. Did the change the defaults for the hsha settings below? How many concurre

Re: Hinted Handoff runs every ten minutes

2012-10-30 Thread Vegard Berget
Hi, I have the exact same problem with 1.1.6.  HintsColumnFamily consists of one row (Rowkey 00, nothing more).   The "problem" started after upgrading from 1.1.4 to 1.1.6.  Every ten minutes HintedHandoffManager starts and finishes  after sending "0 rows".   .vegard, - Original Message -

Re: Hinted Handoff runs every ten minutes

2012-10-30 Thread Vegard Berget
Hi, I have the exact same problem with 1.1.6.  HintsColumnFamily consists of one row (Rowkey 00, nothing more).   The "problem" started after upgrading from 1.1.4 to 1.1.6.  Every ten minutes HintedHandoffManager starts and finishes  after sending "0 rows".   .vegard, - Original Message -

Re: CQL3: Unknown property 'comparator'?

2012-10-30 Thread aaron morton
> My use case relied on "mangling" the column names for various Can you provide details on the use case? CQL 3 is not locked, the more feedback the better. > That's not a problem, but I still needed a definitive statement that > "schema-free tables" (not column families obviously) are now > impo

Re: Multiple counters value after restart

2012-10-30 Thread Jason Wee
maybe enable the debug in log4j-server.properties and going through the log to see what actually happen? On Tue, Oct 30, 2012 at 7:31 PM, Alain RODRIGUEZ wrote: > Hi, > > I have an issue with counters, yesterday I had a lot of ununderstandable > reads/sec on one server. I finally restart Cassand

Re: CQL3 Cell-Transposition - Client or Server Side?

2012-10-30 Thread Sylvain Lebresne
On Tue, Oct 30, 2012 at 11:56 AM, Timmy Turner wrote: > Does the cell transposition that is necessary for CQL3 happen on the > server side after the query execution, or is it something that the > Cassandra/CQL-client does before ultimately handing over the result > set to the caller? The former (

Re: Benifits by adding nodes to the cluster

2012-10-30 Thread Hiller, Dean
1. High availability 2. You can hold much much more data 3. Better performance 4. You can do disaster recovery live-live datacenters (if you configure cassandra) On 10/29/12 4:02 PM, "Andrey Ilinykh" wrote: >This is how cassandra scales. More nodes means better performance. > >thank you, > And

Re: ColumnFamilyInputFormat - error when column name is UUID

2012-10-30 Thread Hiller, Dean
If you want to avoid the type 1 opaque type uuid's, PlayOrm also does a simple key generator if you do @NoSqlId Private String id; It uses the simple hostname(not mac) like a1, a2, a3 which is the names in our cluster plus a unique timestamp within the node. We do not expose the subdomain tha

Data migration between clusters

2012-10-30 Thread 張 睿
Hi, We have several Cassandra clusters in our department, each for a single application. Now we're considering merge these clusters to a single one, and this single one will serve all applications using cassandra, each with a single keyspace. We tried unifying all cluster names to a same one, but

CQL3 Cell-Transposition - Client or Server Side?

2012-10-30 Thread Timmy Turner
Does the cell transposition that is necessary for CQL3 happen on the server side after the query execution, or is it something that the Cassandra/CQL-client does before ultimately handing over the result set to the caller?

RE: idea drive layout - 4 drives + RAID question

2012-10-30 Thread Pieter Callewaert
We also have 4-disk nodes, and we use the following layout: 2 x OS + Commit in RAID 1 2 x Data disk in RAID 0 This gives us the advantage we never have to reinstall the node when a drive crashes. Kind regards, Pieter From: Ran User [mailto:ranuse...@gmail.com] Sent: dinsdag 30 oktober 2012 4:3

Re: Throughput decreases as latency increases with YCSB

2012-10-30 Thread Terje Marthinussen
Check how many concurrent real requests you have vs size of thread pools. Regards, Terje On 30 Oct 2012, at 13:28, Peter Bailis wrote: >> I'm using YCSB on EC2 with one m1.large instance to drive client load > > To add, I don't believe this is due to YCSB. I've done a fair bit of > client-sid

Re: CQL3: Unknown property 'comparator'?

2012-10-30 Thread Timmy Turner
> Is your use case covered in the article above ? My use case relied on "mangling" the column names for various purposes, and CQL3 with its transposed columns does of course still allow for that, but it means rewriting the part of the application that dealt with CQL so it can handle the new syntax