Re: Count of SliceRange in get_slice seems not to work

Just had a quick look at an 0.7b2 install and it appeared to be working as expected.Here's what I got for a row with 50 super columns, that each have 50 columns. I ran the following get_slice calls .get_slice with no super column specified, count=100returned 50 super columns, each with 50 columns g

Re: Count of SliceRange in get_slice seems not to work

It returned all columns within the range of start and end without regard to the count. the CF is super column family and I send the range of super column names of type Long. (and sub column name was UTF8) I put 2000 super columns in a row, and tried to read the first 50 columns in some range of co

Re: Starting up cassandra 0.7

Bryan added a patch. Can someone please review? Thanks. On Wed, Nov 17, 2010 at 12:10 PM, CassUser CassUser wrote: > We are putting together a patch for that. Also updating the unit test to > capture this case. It wasn't failing as a result of it being backwards. > > > On Wed, Nov 17, 2010 at

Re: casssandra 0.7beta3, pycassa and windows client timestamps

No, there's no client side caching. Assuming you're using insert(), could you capture both of these: - the return value of insert() - the same thing you captured before in the Cassandra log so that we can verify that the timestamp is greater than what's stored in Cassandra? - Tyler On Wed, No

RE: casssandra 0.7beta3, pycassa and windows client timestamps

Oh - that was unexpected! I am using pycassa 0.5.2 and thrift 0.5.0 I wasn't aware any client side caching would take place? From: Tyler Hobbs [mailto:ty...@riptano.com] Sent: 17 November 2010 23:34 To: user@cassandra.apache.org Subject: Re: casssandra 0.7beta3, pycassa and windows client timest

Re: casssandra 0.7beta3, pycassa and windows client timestamps

Well, the write appears to be succeeding, and since you say this works on Linux, it sounds like a client side problem. Are you using pycassa or are you dealing with raw Thrift. If pycassa, what version? - Tyler On Wed, Nov 17, 2010 at 5:02 PM, Dr. Andrew Perella wrote: > Thanks Tyler – here i

RE: casssandra 0.7beta3, pycassa and windows client timestamps

Thanks Tyler - here is the log when I call this: DEBUG 22:59:42,914 get_slice DEBUG 22:59:42,915 strongread reading data for SliceFromReadCommand(table='ACRKeyspace_purify', key='31', column_parent='QueryPath(columnFamilyName='UserOwnedCars', superColumnName='null', columnName='null')', start=

Re: casssandra 0.7beta3, pycassa and windows client timestamps

Can you turn on debug logging on Cassandra (change INFO to debug in conf/log4j-server.properties) and see what happens when you send a write that doesn't seem to have an effect? - Tyler On Wed, Nov 17, 2010 at 4:40 PM, Dr. Andrew Perella wrote: > The value I want to change is: > > > > ColumnOrS

RE: casssandra 0.7beta3, pycassa and windows client timestamps

The value I want to change is: ColumnOrSuperColumn(column=Column(timestamp=1290033433618000L, name='currentCarUID', value='82b50f91663f42ee8a740447efa1bdae', ttl=None), super_column=None) and I try to overwrite it with: Mutation(column_or_supercolumn= ColumnOrSuperColumn(column=Column(timestam

Re: casssandra 0.7beta3, pycassa and windows client timestamps

Can you give me an example of what your writes look like here? Are you running a multinode cluster or just one instance of Cassandra? - Tyler On Wed, Nov 17, 2010 at 4:30 PM, Dr. Andrew Perella wrote: > Miliseconds – but that’s not the problem – I can wait 10 minutes and > still not get the va

RE: casssandra 0.7beta3, pycassa and windows client timestamps

Miliseconds - but that's not the problem - I can wait 10 minutes and still not get the value updated! Other times I can update many times a second - it just seems random! From: Tyler Hobbs [mailto:ty...@riptano.com] Sent: 17 November 2010 22:28 To: user@cassandra.apache.org Subject: Re: casssan

Re: casssandra 0.7beta3, pycassa and windows client timestamps

What resolution does time.time() give you in Windows? - Tyler On Wed, Nov 17, 2010 at 4:23 PM, Dr. Andrew Perella wrote: > I have encountered a strange problem with values not being written to > Cassandra very often. > > It looks like it is a windows client timestamp problem. I cannot detect a

casssandra 0.7beta3, pycassa and windows client timestamps

I have encountered a strange problem with values not being written to Cassandra very often. It looks like it is a windows client timestamp problem. I cannot detect a pattern as to when a timestamp will allow an update however if I run my client (in a web server) on linux then the problem goes aw

Re: Rows missing after new node bootstrapped

On Tue, Nov 16, 2010 at 6:58 PM, Jonathan Ellis wrote: > I'm pretty sure that "reading an index" and "using pig" are not > compatible right now. the m/r support that pig builds on always does > sequential-scan range queries. > Yes it does, I have a specialized LoadFunc to read and load manually

Re: using SimpleAuthenticator is not working

One other thing, while login() hangs when called on the server side, and nothing is logged, I see the following logged when I kill the hanged process. DEBUG 13:01:10,640 logged out: null Thanks On 11/17/2010 12:41 PM, Alaa Zubaidi wrote: Hi Aaron, I used the client, and was able to login. E

Re: using SimpleAuthenticator is not working

Hi Aaron, I used the client, and was able to login. E:\cassandra>bin\cassandra-cli.bat -host 191.111.1.11 -port 9160 Starting Cassandra Client Connected to: "Test Cluster" on 191.111.1.11/9160 Welcome to cassandra CLI. Type 'help' or '?' for help. Type 'quit' or 'exit' to quit. [defa...@unknown]

Re: Starting up cassandra 0.7

We are putting together a patch for that. Also updating the unit test to capture this case. It wasn't failing as a result of it being backwards. On Wed, Nov 17, 2010 at 11:51 AM, CassUser CassUser wrote: > Thanks for looking into that. I created a jira issue for it: > > https://issues.apache.or

Re: OPP and controlling partitioning

1) Don't have anything to hand, other than to say the tokens for the OPP are the keys themselves. If your tokens are UUID are the sequential (version 1) UUIDs ? If so you can create a UUID for a known time, with the non type bytes all set high and use that as a token value. Otherwise you will need

Re: Starting up cassandra 0.7

Thanks for looking into that. I created a jira issue for it: https://issues.apache.org/jira/browse/CASSANDRA-1753 On Wed, Nov 17, 2010 at 11:28 AM, Aaron Morton wrote: > That does look like a bug > http://www.mail-archive.com/d...@cassandra.apache.org/msg01167.html > >

Re: Efficienct RangeSlice of SuperColumn and Columns in one call

Duh, forgot that. ThanksAaronOn 18 Nov, 2010,at 01:18 AM, Jonathan Ellis wrote:On Tue, Nov 16, 2010 at 9:08 PM, Aaron Morton wrote: > Super columns have some > limitations http://wiki.apache.org/cassandra/CassandraLimitations > it's not possible to bring back a range of super columns, as you say

Re: Starting up cassandra 0.7

That does look like a bug http://www.mail-archive.com/d...@cassandra.apache.org/msg01167.htmlAaronOn 17 Nov, 2010,at 03:55 PM, Aaron Morton wrote:It's certainly looks suspect. I've had a look at the code around SSTableImport and SSTableExport and the isDeleted value for the col is based on IColumn

Re: Question about load balancing.

Random partitioner distributes keys approximately evenly across the entire range of the ring (0-2**127-1). This means that generally a given section of the range will contain about the same number of keys. If you assign tokens equal-size ranges, they will have similar numbers of keys. This is wh

Re: avro + cassandra + ruby

Full list of client options and defaults: https://github.com/fauna/thrift_client/blob/master/lib/thrift_client/abstract_thrift_client.rb#L28-43 On Wed, Nov 17, 2010 at 10:13 AM, Benjamin Black wrote: > Cassandra.new(keyspace, server, {:protocol => > Thrift::BinaryProtocolAccelerated}) > > On Tue,

Re: avro + cassandra + ruby

Cassandra.new(keyspace, server, {:protocol => Thrift::BinaryProtocolAccelerated}) On Tue, Nov 16, 2010 at 5:13 PM, Ryan King wrote: > On Tue, Nov 16, 2010 at 10:25 AM, Jonathan Ellis wrote: >> On Tue, Sep 28, 2010 at 6:35 PM, Ryan King wrote: >>> One thing you should try is to make thrift use >

Re: OPP and controlling partitioning

@Adi: Yes, that's exactly the reason for the OPP in the Subject :) @Aaron: Thanks for the complete answer. 1) In my case "vertexid_" is a uuid. Could you send me some reference on how to achieve this partitioning based on this prefix and orderpreservingpartitioning? I can't find docs about it.

Re: Smart Client Routing

This is what the StorageProxy API does. There is an example in contrib/client_only. There are some fairly strong limitations: - Java-only - No compatibility guarantees from version to version - You need a separate IP for each client On Tue, Nov 16, 2010 at 9:17 PM, Aaron Morton wrote: > No

Re: Efficienct RangeSlice of SuperColumn and Columns in one call

On Tue, Nov 16, 2010 at 9:08 PM, Aaron Morton wrote: > Super columns have some > limitations http://wiki.apache.org/cassandra/CassandraLimitations > it's not possible to bring back a range of super columns, as you say you can > specify one or zero super column names in your ColumnParent. You spec

Re: Strange values returned

You haven't told Cassandra to treat your columns as anything but bytes. If you want it to know that email is a string, you need to write create column family User with comparator=UTF8Type and column_metadata=[{column_name:email, validation_class:UTF8Type}] On Wed, Nov 17, 2010 at 6:04 AM, Antoni

Strange values returned

Hi from Spain: I'm just doing some samples and I'm obtaining long values instead of the correct value This tests were done in an ubuntu server 10.10 (x86_64) with cassandra 0.7 beta 3 and oracle java 1.6.0_22 What am I doing wrong? Thanks. [defa...@mykeyspace] aserr...@ubuntu:/opt/ cassandra$

Re: Count of SliceRange in get_slice seems not to work

The CassandraServer is not doing the read, step through the code from the call to readColumnFamily() in getSlice(). The read is passed to the StorageProxy.readProtocol() which looks at the CL and determines if its a weak or strong read, sends it out to all the replicas and manages everything. E

Count of SliceRange in get_slice seems not to work

Hello. I'm using cassandra (currently 0.7.0-beta3) in JAVA; with library hector. It seems that cassandra ignore the count of SliceRange when received get_slice request. I traced cassandra source code, and the part of code that retrieving columns does not get count as parameter. See, getSlice