Re: 5 node cluster - Recommended seed configuration.

2011-08-08 Thread Indranath Ghosh
Hi, Rule of thumb, you should identify two servers in the cluster to be your seed nodes. And all the servers should point to these two seeds.. You should bring up the seed nodes first with autobootstrap set to false.. All other nodes should be added to the ring with autobootstrap=true and pointin

Enormous counter problem?

2011-08-08 Thread Boris Yen
Hi, I am not sure if this is a bug or we use the counter the wrong way, but I keep getting a enormous counter number in our deployment. After a few tries, I am finally able to reproduce it. The following are the settings of my development: - I ha

Re: 5 node cluster - Recommended seed configuration.

2011-08-08 Thread samal
It is recommended that seed list should be same in all server so all server on same state. It should be Lan IP not Loopback IP. In seed node, auto bootstrap should be false 2 seed should be enough. In your case it should be like: node1: seeds: node1, autobootstrap=false node2: seeds: node1

SOLVED: Bind the JMX port to a specific IP address/interface

2011-08-08 Thread Teijo Holzer
Hi, I was following this blog about running multiple nodes on the same host: http://www.onemanclapping.org/2010/03/running-multiple-cassandra-nodes-on.html The main nuisance here was that there is no easy way to specify the host IP address for JMX to bind to, it would always bind to all interf

5 node cluster - Recommended seed configuration.

2011-08-08 Thread Selva Kumar
We have a 5 node Cassandra cluster. We use version 0.7.4. What is the recommended seed configuration. Here are some configurations, i have noticed. Example 1: --- One node being seed to itself. node1:     seeds: node1, autobootstrap=false node2:     seeds: node1, node3, autobootstrap=

Re: Trying to find the problem with a broken pipe

2011-08-08 Thread Anthony Ikeda
Tim do you know if this is the actual reason that is causing the broken pipe? I'm having a hard time convincing my team that modifying this value will fix the issue. Jonathan, do you know if there is a valid explanation on why Tim no longer has the problem based on this change? Anthony Ikeda O

[RELEASE] Apache Cassandra 0.8.3 released

2011-08-08 Thread Sylvain Lebresne
The Cassandra team is pleased to announce the release of Apache Cassandra version 0.8.3. Cassandra is a highly scalable second-generation distributed database, bringing together Dynamo's fully distributed design and Bigtable's ColumnFamily-based data model. You can read more here: http://cassand

Re: Chronologic

2011-08-08 Thread Jonathan Ellis
Cool! Do post the slides when you're done. On Mon, Aug 8, 2011 at 8:54 AM, Adam Keys wrote: > I've been working on Chronologic for the past several months. I'm doing a > presentation on it this weekend at Lone Star Ruby Conference. It will be > open source by Sunday 8/14 :). > More info on the t

[update] Cannot insert when RF>=2

2011-08-08 Thread Bret Palsson
Sorry about that sent the last email a little prematurely. -- Does anyone have a solution or have seen this type of behavior? Running: apache-cassandra08-0.8.2 Attempting to use QUORUM reads and writes only works on a keyspace with = RF =3D 1. With RF set to 2 or 3, the write fails and reads

Re: Cannot insert when RF>=2

2011-08-08 Thread Jonathan Ellis
CLI defaults to NTS. You need to specify SimpleStrategy if that's what you want. https://issues.apache.org/jira/browse/CASSANDRA-2960 On Mon, Aug 8, 2011 at 3:10 PM, Bret Palsson wrote: > Running: apache-cassandra08-0.8.2 > > Attempting to use QUORUM reads and writes only works on a keyspace wi

Cannot insert when RF>=2

2011-08-08 Thread Bret Palsson
Running: apache-cassandra08-0.8.2 Attempting to use QUORUM reads and writes only works on a keyspace with RF = 1. With RF set to 2 or 3, the write fails and reads return unavailable. These CLI commands will reproduce the issue: create keyspace rftest with strategy_options=[{replication_factor:2

Re: Chronologic

2011-08-08 Thread Sal Fuentes
This is great news. Thank you Adam and team! I look forward to this gem ... pun intended. :) On Mon, Aug 8, 2011 at 6:54 AM, Adam Keys wrote: > I've been working on Chronologic for the past several months. I'm doing a > presentation on it this weekend at Lone Star Ruby Conference. It will be > o

how to verify the row key is evenly distributed

2011-08-08 Thread myreasoner
Hi all, I have a CF using incremental integer as row keys. In a 5-node cluster with RandomPartitioner, I've noticed the rows are not assigned evenly across nodes--two of them are 5 times heavier loaded than the rest. In nodetool, I can do *getendpoints- Print the end points that owns the k

Re: Replicate on write stage errors

2011-08-08 Thread Philippe
thanks sylvain. I'd ended up finding that issue. and you answered my subsequent questions. Merci Le 8 août 2011 19:32, "Sylvain Lebresne" a écrit :

Re: Replicate on write stage errors

2011-08-08 Thread Sylvain Lebresne
On Sun, Aug 7, 2011 at 11:27 PM, Philippe wrote: > hello, > I've got new errors showing up in my cassandra log file since I starting > testing batch mutates (and it failed). > I have done a rolling restart and they are not disappearing. > How can I fix this ? It's a known bug: https://issues.apac

Re: Setup Cassandra0.8 in Eclipse

2011-08-08 Thread Yang
here is what I normally do: I do not use the ant-generated eclipse files from the build.xml, because that uses external ant build, and every time you make some small code changes, the ant job is invoked, which is annoying. I just create an eclipse java project, select the java sources to be src/j

Re: Setup Cassandra0.8 in Eclipse

2011-08-08 Thread Alvin UW
Thanks guys. After installing svn, it works. But when I followed the last step "If Eclipse still complains about compile errors it is because 'src' (and not 'src/java') is added as source folder. To fix this I recommend to remove 'src' from build path and add 'src/java' to the build path (just lik

Re: batch mutates & throughput

2011-08-08 Thread Philippe
> > Hmm, can you create a ticket with a simple way to reproduce that? We > should be giving back an InvalidRequestException for > multiple-mutations-on-same-key instead of erroring out later and > causing timeouts. > Humm... this is actually quite confusing. When I look at the error, I don't see t

Kundera 2.0.3 Released

2011-08-08 Thread Amresh Singh
We are happy to announce release of Kundera 2.0.3 Kundera is a JPA 2.0 based, Object-Datastore Mapping Library for NoSQL Datastores. The idea behind Kundera is to make working with NoSQL Databases drop-dead simple and fun. It currently supports Cassandra, HBase and MongoDB. Cassandra 0.8 support

Re: batch mutates & throughput

2011-08-08 Thread Philippe
> > Hmm, can you create a ticket with a simple way to reproduce that? We > should be giving back an InvalidRequestException for > multiple-mutations-on-same-key instead of erroring out later and > causing timeouts. > Will do but I'm supposed to be on vacation for a couple weeks now so it will take

Re: Chronologic

2011-08-08 Thread Adam Keys
I've been working on Chronologic for the past several months. I'm doing a presentation on it this weekend at Lone Star Ruby Conference. It will be open source by Sunday 8/14 :). More info on the talk: http://speakerrate.com/talks/7808 On Aug 6, 2011, at 3:57 AM, Sal Fuentes wrote: > Hello Scot

Re: Setup Cassandra0.8 in Eclipse

2011-08-08 Thread Jake Luciani
you can simply run: ant generate-eclipse-files then import the project On Sun, Aug 7, 2011 at 5:39 PM, Alvin UW wrote: > Hello, > > I am trying to Setup Cassandra0.8 in Eclipse following > http://wiki.apache.org/cassandra/RunningCassandraInEclipse > After right clicking on the build.xml -> "Ru

Re: batch mutates & throughput

2011-08-08 Thread Jonathan Ellis
Hmm, can you create a ticket with a simple way to reproduce that? We should be giving back an InvalidRequestException for multiple-mutations-on-same-key instead of erroring out later and causing timeouts. On Mon, Aug 8, 2011 at 12:34 AM, Philippe wrote: > Hi Boris, > Thanks for the suggestion, I

per node meta data & configuration

2011-08-08 Thread Edward Capriolo
When cassandra loaded from the xml file it was easy to experiment with different settings. With new row cache providers and the new sstable compression there is a problem. While these can be tested for correctness on staging gear. In production these things have to be rolled out all nodes or none.

Re: Last manual repair time

2011-08-08 Thread Bart Swedrowski
On 4 August 2011 22:59, Teijo Holzer wrote: > End of every repair: > grep 'No neighbors to repair with' system.log Thanks - very helpful. Already wrote a tiny nagios plugin to go through this and alert when needed.