cassandra-user mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From Jonathan Ellis <jbel...@gmail.com>
Subject Re: writes to Cassandra failing occasionally
Date Thu, 08 Apr 2010 15:56:55 GMT
is N:C:E possibly ignoring thrift exceptions?

On Thu, Apr 8, 2010 at 10:45 AM, Mike Gallamore
<mike.e.gallamore@googlemail.com> wrote:
> On 04/08/2010 04:53 AM, Philip Jackson wrote:
>>
>> At Wed, 07 Apr 2010 13:19:26 -0700,
>> Mike Gallamore wrote:
>>
>>>
>>> I have writes to cassandra that are failing, or at least a read shortly
>>> after a write is still getting an old value. I realize Cassandra is
>>> "eventually consistent" but this system is a single CPU single node with
>>> consistency level set to 1, so this seems odd to me.
>>>
>>
>> I'm having this problem too (see my post the other day). I use N::C
>> but generate timestamps in the same way as N::C::E, I've tested that
>> each is smaller than the next so I'm wondering if I'm barking up the
>> wrong tree.
>>
>> If you figure out what's going on please do post back here, I'll do
>> the same.
>>
>> Cheers,
>> Phil
>>
>
> I modified N::C::E to output the timestamp each time the timestamp
> subroutine is called. I've confirmed that the timestamp is increasing (never
> getting two updates with the same timestamp) but I'm still loosing data.
>
> This is painful because part of the data structure I'm working with is a
> revolving one week bucket of daily scores. When the system forgets that the
> day has changed it thinks it is still yesterday and "nukes" the data in that
> bucket (which it just set) thinking that it is week old data. The desired
> behavior of course is that it realize that it is a new day (this is stored
> in the data structure itself and isn't modifible as it depends on when the
> data was first created for a particular row which "bucket" the system will
> be working with any given day) and nuke the new day's old data before
> updating it with a new score.
>

Mime
View raw message