Re: Cassandra crashed - possible JMX threads leak

Ah, missed that. Thanks Aaron. On Wed, Oct 20, 2010 at 11:49 PM, Aaron Morton wrote: > Sounds like the problem discussed here > http://wiki.apache.org/cassandra/Operations?highlight=(allocate)|(memory) > > >

Re: Cassandra crashed - possible JMX threads leak

Sounds like the problem discussed here http://wiki.apache.org/cassandra/Operations?highlight=(allocate)|(memory)If you have the JNA jar it should work http://www.mail-archive.com/user@cassandra.apache.org/msg06284.htmlhttp://www.riptano.com/blog/whats-new-cassandra-065AaronOn 21 Oct, 2010,at 06:29

Re: Cassandra crashed - possible JMX threads leak

I have a cluster of 3 0.7 beta 2 nodes (built today from the latest trunk) running on Large, EBS-backed, x64 EC2 instances; RF=3. I attempted to write somewhere near 500,000 records every 15 minutes from a total of 5 different computers (using Pelops and multi-threading). Though my network blew

Re: What happens if there is a collision?

On 21/10/2010 4:43, Jonathan Ellis wrote: On Wed, Oct 20, 2010 at 8:38 PM, Jérôme Verstrynge wrote: -) Let's imagine node A updates MyColumn with value 'AAA' and timestamp 3567890 -) Let's imagine node E updates MyColumn with value 'EEE' and timestamp 3567890 What happens? Who wins? Is it dete

Re: What happens if there is a collision?

On Wed, Oct 20, 2010 at 8:38 PM, Jérôme Verstrynge wrote: > -) Let's imagine node A updates MyColumn with value 'AAA' and timestamp > 3567890 > -) Let's imagine node E updates MyColumn with value 'EEE' and timestamp > 3567890 > > What happens? Who wins? Is it deterministic? value breaks ties if t

What happens if there is a collision?

Hi, I am a new to Cassandra. I am reading all the documentation I can find online. My question is the following: -) Let's imagine a cluster with 5 nodes ABCDE. We know that quorum = 3. -) Let's imagine a column called MyColumn. -) Let's imagine current timestamp = 3567890. -) Let's imagine no

Tried to add a new node but f*ed up

I tried adding a new node and rebalanced the ring via nodetool move but ending up in a weird state. Blew away all data from 2 nodes (out of 3) and manually set tokens but its completely unbalanced. [r...@cassandra1 apache-cassandra]# bin/nodetool --host localhost --port 8080 ring Address

Re: Transaction/Locking Support

There are no transactions or locking in cassandra, if you want them you need to use Zookeeper or similar. Writes for a single mutation are atomic but  not isolated http://wiki.apache.org/cassandra/FAQ#batch_mutate_atomic More about consistency here http://wiki.apache.org/cassandra/ArchitectureOverv

Transaction/Locking Support

Hi, I'm trying to persist the contents of my app. objects to Cassandra. The objects in here are loosely modeled on key value (KV) stores. If the KV store is extended from HashMap and mapped to a ColumnFamily with put and get ops mapping to insert/read from Cassandra. Then my questions are: a) What

Re: Throttling ColumnFamilyRecordReader

sorry i had a misunderstanding of the MapRed report output. i did reduce mapreduce.tasktracker.map.tasks.maximum (number of concurrent maps per node) from the default of 2 to 1. i suppose if i want to do this on a per job/user basis i'll try out the hadoop fair scheduler. On Oct 19, 2010, at

Re: Hector question under cassandra 0.7

I figure I'd reply to my own question in case this helps others. Talking on the IRC, having one KeyspaceOperator per thread (via ThreadLocal) makes sense. On Wed, Oct 20, 2010 at 9:13 AM, Ned Wolpert wrote: > Folks- > > I'm finally upgrading the grails-cassandra plugin for 0.7, and wanted to >

Re: memtable sstable questions (0.6.4)

For capacity planning it's not worth worrying about whether the MemTables are empty, they will all end up full.Internal caches may refer to either the Row and Key caches or the BloomFilters, not sure which in this case. AaronOn 21 Oct, 2010,at 09:42 AM, CassUser CassUser wrote:I didn't notice the

Re: Cassandra crashed - possible JMX threads leak

can you reproduce this by, say, running nodeprobe ring in a bash while loop? On Wed, Oct 20, 2010 at 3:09 PM, Bill Au wrote: > One of my Cassandra server crashed with the following: > > ERROR [ACCEPT-xxx.xxx.xxx/nnn.nnn.nnn.nnn] 2010-10-19 00:25:10,419 > CassandraDaemon.java (line 82) Uncaught ex

Re: memtable sstable questions (0.6.4)

I didn't notice the number of hot CFs mentioned below. So with data in them. We are sharing a cluster with others, so I'm trying to get an idea of what overhead there is for empty CFs if any. What are internal caches? On Wed, Oct 20, 2010 at 1:17 PM, CassUser CassUser wrote: > Cool thanks, that

Re: memtable sstable questions (0.6.4)

Cool thanks, that helps. So even if we have defined a column family in the storage-conf and it's empty, this has some overhead in cassandra and the following rule should apply: memtable_throughput_in_mb * 3 * number of hot CFs + 1G + internal caches. On Wed, Oct 20, 2010 at 12:53 PM, Aaron Mor

Cassandra crashed - possible JMX threads leak

One of my Cassandra server crashed with the following: ERROR [ACCEPT-xxx.xxx.xxx/nnn.nnn.nnn.nnn] 2010-10-19 00:25:10,419 CassandraDaemon.java (line 82) Uncaught exception in thread Thread[ACCEPT-xxx.xxx.xxx/nnn.nnn.nnn.nnn,5,main] java.lang.OutOfMemoryError: unable to create new native thread

Re: memtable sstable questions (0.6.4)

Take a look at the section on JVM Heap size here http://wiki.apache.org/cassandra/MemtableThresholdsCF's have a large overhead, Keyspaces have none/little. In general write performance will be affected by the memtable thresholds (also on the link above). Read performance will be affected by the siz

Re: memtable sstable questions (0.6.4)

Thanks for the link. #2 was not meant to be trick question, it just came out like that :). what i was after is the overhead associated with large number of keyspaces and column families (i didn't mean empty memtables :). If a few keyspaces that have 20 or so column families with a percentage of

Re: Read Latency

If you are using Python, and raw Thrift, use the following: protocol = TBinaryProtocol.TBinaryProtocolAccelerated(transport) The serialization/deserialization is done directly in C. On Wed, Oct 20, 2010 at 11:53 AM, Wayne wrote: > We did some testing and the object is 23megs that is taking mor

Re: Hadoop Word Count Super Column Example?

We're the IColumn objects passed to the map function o.a.c.db.SuperColumn instances ? AOn 21 Oct, 2010,at 02:48 AM, Jeremy Hanna wrote:> Have your tried it ? yes, with a modified word count example a month or so ago. On Oct 20, 2010, at 3:27 AM, aaron morton wrote: > My understanding of the Ha

Re: memtable sstable questions (0.6.4)

On Wed, Oct 20, 2010 at 2:47 PM, CassUser CassUser wrote: > Hey, > > As I understand it writes go directly to the commit log.  Once a threshold > has been reached the data is shipped to a memtable, and again to an sstable. > > 1. How many memtables are created when a flush happens from a commit lo

Re: Read Latency

We did some testing and the object is 23megs that is taking more than 3 seconds for thrift to return as a python object. We also tested pickling this object to/from a string and to pickle it takes 1.5s and to convert the pickled string to a python object takes .75s. Added together they still take l

memtable sstable questions (0.6.4)

Hey, As I understand it writes go directly to the commit log. Once a threshold has been reached the data is shipped to a memtable, and again to an sstable. 1. How many memtables are created when a flush happens from a commit log? One per CF? 2. Is there any space associated with an empty memtab

Re: Client listener for Cassandra Column Family Updates

Short answer: no. On Wed, Oct 20, 2010 at 10:15 AM, Potter,Lorne [Wpg] wrote: > I have a 0.7beta2 four node cluster set up and a small Java program that > writes to a TimeUUIDType sorted column family and another program that polls > the database every x msecs to read the latest data. Instead of

Re: Client listener for Cassandra Column Family Updates

On Wed, Oct 20, 2010 at 11:15 AM, Potter,Lorne [Wpg] wrote: > I have a 0.7beta2 four node cluster set up and a small Java program that > writes to a TimeUUIDType sorted column family and another program that polls > the database every x msecs to read the latest data. Instead of polling I > would r

Re: Preventing an update of a CF row

> Always specify some constant value for timestamp. Only 1st insertion with that > timestamp will succeed. Others will be ignored, because will be considered > duplicates by cassandra. Just be aware that this limits you in several ways in terms of your data model. For examle, if you ever need to r

Re: How fast does compaction run?

> We are using 0.6.1, and are planning to upgrade to 0.6.6.  Currently, we are > observing that the compaction is going at the rate about ~10MB per second. > We did not see obvious bottleneck hardware wise. > > > > Is this normal, or did we miss something? Assuming you're otherwise not entirely sa

Re: Cassandra CF sharding

Hi Maxim, Curious to know how your experimentation with CF Sharding go. We have similar limitations and am trying to tackle the exact same problem. Anybody else have any suggestions/experiments that they have tried around this? Thanks Anand -- View this message in context: http://cassandra-us

How fast does compaction run?

We are using 0.6.1, and are planning to upgrade to 0.6.6. Currently, we are observing that the compaction is going at the rate about ~10MB per second. We did not see obvious bottleneck hardware wise. Is this normal, or did we miss something? The information tra

Hector question under cassandra 0.7

Folks- I'm finally upgrading the grails-cassandra plugin for 0.7, and wanted to understand a bit more on the usage of the Cluster and KeyspaceOperator . Is the cluster object retrieved from HFactor.createCluster() thread safe, and is the KeyspaceOperator required to only be in one thread? Or are

Re: Authentication

Hello Nate, Thank you very much for the reply. Next time I will send to hector-us...@googlegroups.com Regards, Pablo D. Salgado 2010/10/20 Nate McCall > See this issue for background: > > http://groups.google.com/group/hector-users/browse_thread/thread/b4c98bbd382cbe56 > > Authentication will

Client listener for Cassandra Column Family Updates

I have a 0.7beta2 four node cluster set up and a small Java program that writes to a TimeUUIDType sorted column family and another program that polls the database every x msecs to read the latest data. Instead of polling I would rather established some kind listener so Cassandra would notify the cl

Re: Authentication

See this issue for background: http://groups.google.com/group/hector-users/browse_thread/thread/b4c98bbd382cbe56 Authentication will be picked up again shortly given the pending 0.7.0 release. Also, please feel free to send Hector-specific questions to: hector-us...@googlegroups.com On Wed, Oct

Authentication

Hello everybody, I'm working with Hector 0.6.0_12 and Cassandra 0.6.6 with SimpleAuthenticator. How could I login to Cassandra using Hector? - With the standard operation (without sending login information to Hector) I'm getting this exception: INFO me.prettyprint.cassandra.service.CassandraClie

Re: Hadoop Word Count Super Column Example?

> Have your tried it ? yes, with a modified word count example a month or so ago. On Oct 20, 2010, at 3:27 AM, aaron morton wrote: > My understanding of the Hadoop integration is not great but from what I can > see. The code in o.a.c.hadoop.ColumnFamilyRecordReader does not use a > super_colu

Re: Hadoop Word Count Super Column Example?

My understanding of the Hadoop integration is not great but from what I can see. The code in o.a.c.hadoop.ColumnFamilyRecordReader does not use a super_column in the ColumnParent struct when making the get_range_slices() call. It's just using the ColumnFamily. So I would guess it would includ