incubator-cassandra-user mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From Michal Michalski <mich...@opera.com>
Subject Re: nodetool ring showing different 'Load' size
Date Wed, 19 Jun 2013 12:47:18 GMT
You can start compaction via JMX if you need it and you know what you're 
doing:
Find org.apache.cassandra.db:type=CompactionManager MBean and 
forceUserDefinedCompaction operation in it.
First argument is keyspace name, second one is a comma-separated list of 
SSTables to compact (filename)

You can also perform a major compaction via nodetool compact (for 
SizeTieredCompaction), but - again - you really should not do it unless 
you're really sure what you do, as it compacts all the SSTables 
together, which is not something you might want to achieve in most of 
the cases.

M.

W dniu 19.06.2013 14:31, Rodrigo Felix pisze:
> Thanks Eric. Is there a way to start manually compaction operations?
> I'm thinking about doing after loading data and before start run phase of
> the benchmark.
> Thanks.
>
> Att.
>
> *Rodrigo Felix de Almeida*
> LSBD - Universidade Federal do CearĂ¡
> Project Manager
> MBA, CSM, CSPO, SCJP
>
>
> On Mon, Jun 17, 2013 at 12:41 PM, Eric Stevens <mightye@gmail.com> wrote:
>
>> Load is the size of the storage on disk as I understand it.  This can
>> fluctuate during normal usage even if records are not being added or
>> removed, a node's load may be reduced during compaction for example.
>>   During compaction, especially if you use Size Tiered Compaction strategy
>> (the default), load may temporarily double for a column family.
>>
>>
>> On Mon, Jun 17, 2013 at 11:33 AM, Rodrigo Felix <
>> rodrigofelixdealmeida@gmail.com> wrote:
>>
>>> Hi,
>>>
>>>     I've been running a benchmark on Cassandra and I'm facing a problem
>>> regarding to the size of the database.
>>>     I performed a load phase and then, when running nodetool ring, I got
>>> the following output:
>>>
>>> *ubuntu@domU-12-31-39-0E-11-F1:~/cassandra$ bin/nodetool ring *
>>> *Address         DC          Rack        Status State   Load
>>>   Effective-Ownership Token                                       *
>>> *
>>>                     85070591730234615865843651857942052864      *
>>> *10.192.18.3     datacenter1 rack1       Up     Normal  2.07 GB
>>> 50.00%              0                                           *
>>> *10.85.135.169   datacenter1 rack1       Up     Normal  2.09 GB
>>> 50.00%              85070591730234615865843651857942052864*
>>>
>>>     After that I executed, for about one hour, a workload with scan and
>>> insert queries. Then, after finishing the workload execution, I run again
>>> nodetool ring and got the following:
>>>
>>> *ubuntu@domU-12-31-39-0E-11-F1:~/cassandra$ bin/nodetool ring *
>>> *Address         DC          Rack        Status State   Load
>>>   Effective-Ownership Token                                       *
>>> *
>>>                     85070591730234615865843651857942052864      *
>>> *10.192.18.3     datacenter1 rack1       Up     Normal  1.07 GB
>>> 50.00%              0                                           *
>>> *10.85.135.169   datacenter1 rack1       Up     Normal  2.15 GB
>>> 50.00%              85070591730234615865843651857942052864*
>>>
>>>     Any idea why a node had its size reduced if no record was removed? No
>>> machine or added or removed during this workload.
>>>     Is this related to any kind of compression? If yes, is there a command
>>> to confirm that?
>>>     I also faced a problem where a node has its size increased from about
>>> 2gb to about 4gb. In this last scenario, I both added and removed nodes
>>> during the workload depending on the load (CPU).
>>>     Thanks in advance for any help.
>>>
>>>
>>> Att.
>>>
>>> *Rodrigo Felix de Almeida*
>>> LSBD - Universidade Federal do CearĂ¡
>>> Project Manager
>>> MBA, CSM, CSPO, SCJP
>>>
>>
>>
>


Mime
View raw message