Re: Cassandra OOM crash while mapping commitlog

2012-08-13 Thread Robin Verlangen
@Tyler: We were already running most of our machines in 64bit JVM (Sun, not the OpenJDK). Those also crashed. @Holger: Good to hear that. I'll schedule an update for our Cassandra cluster. Thank you both for your time. 2012/8/13 Holger Hoffstaette holger.hoffstae...@googlemail.com On Sun, 12

Re: Loading data on-demand in Cassandra

2012-08-13 Thread Pierre Chalamet
Hi, You do not need to have 700 Go of data in RAM. Cassandra is able to store on disks and query from there if data is not cached in memory. Caches are maintained by C* by itself but you still have to some configuration. Supposing you want to store around 800 Go and with a RF=3, you will need

Re: [gem] does disconnect! work properly?

2012-08-13 Thread aaron morton
My rough understanding of the ruby code is that it auto connects if necessary when you make a request. i.e. your get turns into a _multiget call here https://github.com/twitter/cassandra/blob/master/lib/cassandra/protocol.rb#L52 and that result is in a client connection if necessary

Re: Problem with cassandra startup on Linux

2012-08-13 Thread aaron morton
Hi Dwight, I can confirm that issue on my MBP under Mountain Lion. Can you create a ticker at https://issues.apache.org/jira/browse/CASSANDRA and include the platform you are running on. For reference the change was added by https://issues.apache.org/jira/browse/CASSANDRA-4447 The

RE: RE Restore snapshot

2012-08-13 Thread mdione.ext
De : Sylvain Lebresne [mailto:sylv...@datastax.com] 2) copy the snapshot sstable in the right place and call the JMX method loadNewSSTables() (in the column family MBean, which mean you need to do that per-CF). How does this affect the contents of the CommitLogs? I mean, I

RE: RE Restore snapshot

2012-08-13 Thread mdione.ext
De : mdione@orange.com [mailto:mdione@orange.com] De : Sylvain Lebresne [mailto:sylv...@datastax.com] 2) copy the snapshot sstable in the right place and call the JMX method loadNewSSTables() (in the column family MBean, which mean you need to do that per-CF). How

Re: DSE solr HA

2012-08-13 Thread Jake Luciani
Going through this page and it looks like indexes are stored locally http://www.datastax.com/dev/blog/cassandra-with-solr-integration-details . My question is what happens if one of the solr nodes crashes? Is the data indexed again on those nodes? Yes the data is indexed again

RE: Problem with cassandra startup on Linux

2012-08-13 Thread Dwight Smith
Aaron Yes will do - I had already made the suggested change - cluster is up and running. Thanks From: aaron morton [mailto:aa...@thelastpickle.com] Sent: Monday, August 13, 2012 1:56 AM To: user@cassandra.apache.org Subject: Re: Problem with cassandra startup on Linux Hi Dwight,

RE: Problem with cassandra startup on Linux

2012-08-13 Thread Dwight Smith
Aaron https://issues.apache.org/jira/browse/CASSANDRA-4535 created. Thanks much From: Dwight Smith [mailto:dwight.sm...@genesyslab.com] Sent: Monday, August 13, 2012 7:54 AM To: user@cassandra.apache.org Subject: RE: Problem with cassandra startup on Linux Aaron Yes will

Re: 1.1.3 crasch when initializing column family

2012-08-13 Thread Omid Aladini
It works for me on Sun's jvm. It appears to be similar to: http://www-01.ibm.com/support/docview.wss?uid=swg1IV12627 Have you tried upgrading IBM's java or using Sun's? -- Omid On Mon, Aug 13, 2012 at 1:34 PM, Robert Hellmans robert.hellm...@aastra.com wrote: ** Hi I'm trying the

Re: Cassandra OOM crash while mapping commitlog

2012-08-13 Thread Robin Verlangen
3 hours ago I finished the upgraded of our cluster. Currently it runs quite smooth. I'll give an update within a week if this really solved our issues. Cheers! 2012/8/13 Robin Verlangen ro...@us2.nl @Tyler: We were already running most of our machines in 64bit JVM (Sun, not the OpenJDK).

Custom Partitioner Type

2012-08-13 Thread A J
Is it possible to use a custom Partitioner type (other than RP or BOP) ? Say if my rowkeys are all Integers and I want all even keys to go to node1 and odd keys to node2, is it feasible ? How would I go about ? Thanks.

Re: Custom Partitioner Type

2012-08-13 Thread aaron morton
Yes, you need to implement the org.apache.cassandra.dht.IPartitioner interface, there are a couple of abstract implementations you could base it on. I want all even keys to go to node1 and odd keys to node2, is it feasible ? I'm not endorsing the idea of doing this, but as a hack to see if

GCInspector info messages in cassandra log

2012-08-13 Thread Tamar Fraenkel
Hi! I have 3 nodes ring running on Amazon EC2. About once a week I see in the logs compaction messages and around the same time info messages about GC (see below) that I think means it is taking too long and happening too often. Does it mean I have to reduce my cache size? Thanks, Tamar INFO

What are the basic steps to improve Cassandra performance

2012-08-13 Thread A Geek
hi all, I'm a bit new to Cassandra and was wondering what are the basic steps that we must follow to optimize the Cassandra for performance in general and how to improve the performance for select * from X type of queries. Any help would be much appreciated. Note that, we have huge data