Tamar Fraenkel | 1 May 06:59 2012
Picon

Re: Hector 1.0-5 released

Thanks, can you post a link to the jar when you do?
Tamar

On Apr 30, 10:01 pm, Nate McCall <zznat...@...> wrote:
> Yeah actually. I'll put one up later this evening. (Thanks for reminding me).
>
>
>
>
>
>
>
> On Mon, Apr 30, 2012 at 1:58 PM, Aaron Turner <synfina...@...> wrote:
> > Is there going to be a 1.0.5 tarball posted to github?
>
> > On Mon, Apr 23, 2012 at 10:51 PM, Nate McCall <zznat...@...> wrote:
> >> Should be winding it's way up to maven central currently.
>
> >> Scrape from the changelog:
> >> ************
> >> *** We moved the repo tohttps://github.com/hector-client/hector
> >> thanks to  <at> rantav for all his hard work and effort over the course of
> >> the
> >> past two years on this project and hosting everything for as long as he did ***
> >> ************
> >> Remove ExhaustedPolicy as it doesn't seem to be used anymore.  This
> >> will probably break some folks' configuration - watch out for spring
> >> config!
> >> Added listener framework for CassandraHostRetry service per issue #455
> >> huge thanks to  <at> elyrank
(Continue reading)

Nate McCall | 1 May 18:08 2012
Picon

Re: Hector 1.0-5 released

https://github.com/downloads/hector-client/hector/hector-core-1.0-5.tgz

On Mon, Apr 30, 2012 at 11:59 PM, Tamar Fraenkel
<tamar.fraenkel@...> wrote:
> Thanks, can you post a link to the jar when you do?
> Tamar
>
> On Apr 30, 10:01 pm, Nate McCall <zznat...@...> wrote:
>> Yeah actually. I'll put one up later this evening. (Thanks for reminding me).
>>
>>
>>
>>
>>
>>
>>
>> On Mon, Apr 30, 2012 at 1:58 PM, Aaron Turner <synfina...@...> wrote:
>> > Is there going to be a 1.0.5 tarball posted to github?
>>
>> > On Mon, Apr 23, 2012 at 10:51 PM, Nate McCall <zznat...@...> wrote:
>> >> Should be winding it's way up to maven central currently.
>>
>> >> Scrape from the changelog:
>> >> ************
>> >> *** We moved the repo tohttps://github.com/hector-client/hector
>> >> thanks to  <at> rantav for all his hard work and effort over the course of
>> >> the
>> >> past two years on this project and hosting everything for as long as he did ***
>> >> ************
>> >> Remove ExhaustedPolicy as it doesn't seem to be used anymore.  This
(Continue reading)

Nate McCall | 1 May 21:41 2012
Picon

1.1 patch set in from Ed C.

Huge thanks to Ed C. for the update patch.

I told him to PR with the few remaining errors, cause we need to
hustle on this. I'll be on this (real) late this evening, but if
anyone else has some cycles and want's to trouble shoot (slice_ranges
w/ super columns started chocking for some reason in ApiV2SystemTest
and VirtualKeyspaceTest [which just extends the former]).

My plan is this:
- create a 1.0 branch in hector
- push 1.0 branch up to master
- begin tracking trunk on 1.1 after pulling in Ed's changes

I'm waiting on the ack from Sonatype for setting up the new repository
coordinates, but should have that in the next 48hrs at the latest.

Mohit Anchlia | 1 May 21:50 2012
Picon

How does slice having start and end with reverse works

If I have column names like:
 
A.1
A.1
B.1
B.2
 
And I have a slice start of "B. + Character.MIN_VALUE" and end of "B. + Character.MAX_VALUE" with reversed set to true and slice.count(1). Does the search start from bottom up and returns the first one it finds?
 
I essentially want to retrieve only B.2 column
David Fischer(Gtalk | 2 May 19:02 2012
Picon

setExhaustedPolicy removal

Nate,

in hector 1.0-5 you removed setExhaustedPolicy.  What is now the
replacement of this?

Nate McCall | 2 May 19:06 2012
Picon

Re: setExhaustedPolicy removal

This setting was not being used and was actually a hold over from
early 0.7 releases. The semantics of the underlying ArrayBlockingQueue
with the thread count cap and acquisition timeout negate the need for
this.

On Wed, May 2, 2012 at 12:02 PM, David Fischer(Gtalk)
<fischer.d.r@...> wrote:
> Nate,
>
> in hector 1.0-5 you removed setExhaustedPolicy.  What is now the
> replacement of this?

Picon

Re: Specifying OR condition in hector

I guess there's no support for OR condition on cql:


http://cassandra.apache.org/doc/cql/CQL.html#SELECT 




On Mon, Apr 30, 2012 at 4:48 PM, Mohit Anchlia <mohitanchlia-Re5JQEeQqe8AvxtiuMwx3w@public.gmane.org> wrote:
I wonder how CQL would do this if it's not supported at Thrift level? Or may CQL doesn't support anOR condition?


On Mon, Apr 30, 2012 at 12:43 PM, Nate McCall <zznate.m-Re5JQEeQqe8AvxtiuMwx3w@public.gmane.org> wrote:
Nope - you effectively have to do a merge-join on the client.

An API wrapping this up at least at the interface definition level
would be pretty slick though.

On Mon, Apr 30, 2012 at 2:25 PM, Mohit Anchlia <mohitanchlia-Re5JQEeQqe8AvxtiuMwx3w@public.gmane.org> wrote:
> Is there a way to query on 2 different slice predicates in one query call?
> for eg: col name A.* and col name B.*
>
> I asked it before but didn't get reply and I don't see anything in hector
> that supports this. Just wondering if this can easily be added?


Picon

Re: Specifying OR condition in hector

BTW, take care with cql statements. Mainly those which doesn't use any condition.





On Wed, May 2, 2012 at 5:15 PM, Víctor Hugo Oliveira Molinar <vhmolinar-Re5JQEeQqe8AvxtiuMwx3w@public.gmane.org> wrote:
I guess there's no support for OR condition on cql:

http://cassandra.apache.org/doc/cql/CQL.html#SELECT 





On Mon, Apr 30, 2012 at 4:48 PM, Mohit Anchlia <mohitanchlia-Re5JQEeQqe8AvxtiuMwx3w@public.gmane.org> wrote:
I wonder how CQL would do this if it's not supported at Thrift level? Or may CQL doesn't support anOR condition?


On Mon, Apr 30, 2012 at 12:43 PM, Nate McCall <zznate.m-Re5JQEeQqe8AvxtiuMwx3w@public.gmane.org> wrote:
Nope - you effectively have to do a merge-join on the client.

An API wrapping this up at least at the interface definition level
would be pretty slick though.

On Mon, Apr 30, 2012 at 2:25 PM, Mohit Anchlia <mohitanchlia-Re5JQEeQqe8AvxtiuMwx3w@public.gmane.org> wrote:
> Is there a way to query on 2 different slice predicates in one query call?
> for eg: col name A.* and col name B.*
>
> I asked it before but didn't get reply and I don't see anything in hector
> that supports this. Just wondering if this can easily be added?



Aaron Turner | 3 May 06:59 2012
Picon

1.0-5 and org/apache/thrift/TEnum ??

I recently updated from Hector 0.8-x to 1.0-5.

after upgrading hector at startup I'm now getting the following error:

NameError: cannot link Java class
org.apache.cassandra.thrift.ConsistencyLevel, probable missing
dependency: org/apache/thrift/TEnum

Now, you're prolly thinking, "that doesn't look like a normal java
error", and that's because it's actually from a JRuby application.  I
dug through the jars which shipped with 1.0-5, and I couldn't find any
class named org.apache.thrift.TEnum, so I'm not sure where it's
supposed to be coming from or what might be the cause of the error.
Other then swapping hector versions, everything else is the same.  Any
ideas?  I read through the changelog and readme and didn't notice
anything that talks about this.

Thanks!

Full backtrace:
NameError: cannot link Java class
org.apache.cassandra.thrift.ConsistencyLevel, probable missing
dependency: org/apache/thrift/TEnum
          for_name at org/jruby/javasupport/JavaClass.java:1205
   get_proxy_class at org/jruby/javasupport/JavaUtilities.java:34
       java_import at
/home/tools/torquebox/jruby/lib/ruby/site_ruby/shared/builtin/javasupport/core_ext/object.rb:46
            (root) at
/home/tools/spider-svn/spider/config/initializers/java.rb:28
              load at org/jruby/RubyKernel.java:1068
              load at
/home/tools/spider-svn/spider/config/initializers/java.rb:245
   load_dependency at
/home/tools/gems/jruby/1.9/gems/activesupport-3.2.3/lib/active_support/dependencies.rb:236
              load at
/home/tools/gems/jruby/1.9/gems/activesupport-3.2.3/lib/active_support/dependencies.rb:245
            Engine at
/home/tools/gems/jruby/1.9/gems/railties-3.2.3/lib/rails/engine.rb:588
              each at org/jruby/RubyArray.java:1615
            Engine at
/home/tools/gems/jruby/1.9/gems/railties-3.2.3/lib/rails/engine.rb:587
     instance_exec at org/jruby/RubyBasicObject.java:1757
               run at
/home/tools/gems/jruby/1.9/gems/railties-3.2.3/lib/rails/initializable.rb:30
  run_initializers at
/home/tools/gems/jruby/1.9/gems/railties-3.2.3/lib/rails/initializable.rb:55
              each at org/jruby/RubyArray.java:1615
  run_initializers at
/home/tools/gems/jruby/1.9/gems/railties-3.2.3/lib/rails/initializable.rb:54
       initialize! at
/home/tools/gems/jruby/1.9/gems/railties-3.2.3/lib/rails/application.rb:136
          __send__ at org/jruby/RubyBasicObject.java:1698
              send at org/jruby/RubyKernel.java:2097
    method_missing at
/home/tools/gems/jruby/1.9/gems/railties-3.2.3/lib/rails/railtie/configurable.rb:30
            (root) at
/home/tools/spider-svn/spider/./script/../config/environment.rb:7
           require at org/jruby/RubyKernel.java:1042
           require at
/home/tools/spider-svn/spider/./script/../config/environment.rb:36
            (root) at ./script/load_ncp.rb:2

--

-- 
Aaron Turner
http://synfin.net/         Twitter:  <at> synfinatic
http://tcpreplay.synfin.net/ - Pcap editing and replay tools for Unix & Windows
Those who would give up essential Liberty, to purchase a little temporary
Safety, deserve neither Liberty nor Safety.
    -- Benjamin Franklin
"carpe diem quam minimum credula postero"

amy | 3 May 15:24 2012

Re: Hector behavior in multiple DC cluster

Thanks Nate,

In our use care, we'd rather the requests got across the WAN than
returns a failed read/write to the client.

We tried DynamicLoadBalancingPolicy, but it was still hitting the
seemingly good node in the bad dc, and returned a failure to the
client.

We are now trying out a different Balancing Policy which injected with
DC details.  We filter out those 'seemingly active node in a bad dc'
then passed the pools to the default LB Policy.


Gmane