Hi Eric,

no, it's not related to Hazelcast, it's just that OrientDB relies on a
backup to a new replica, and now backups are a physical copy of disk
structure. It's just a technical detail, nothing structural, so we expect
to have a fix in short time

Thanks

luigi


2015-07-31 18:01 GMT+02:00 Eric24 <[email protected]>:

> Ah, OK. I didn't realize that data needed to be persisted to disk in order
> to have a distributed cluster (is that a Hazelcast requirement?). I'm also
> glad to hear that this is on the road map--I can see numerous use-cases for
> a reliable (i.e. HA clustered) in-memory database.
> --Eric
>
>
> On Friday, July 31, 2015 at 4:09:21 AM UTC-5, Luigi Dell'Aquila wrote:
>>
>> Well, actually memory storage structure is exactly the same as the
>> persistent one, the only difference is that it does not flush data to disk.
>> Some of our customers use OrientDB in-memory only, also on production
>> instances.
>> You just have to keep in mind that this has some limitations; the biggest
>> one is that you cannot have a distributed cluster of all in-memory
>> instances, just because node replica is based on backup procedure, that is
>> executed on physical files (so it cannot be executed on in-memory db). So
>> if you want to try a distributed configuration, at least one node has to be
>> a persistent one. This limitation will be removed in next releases.
>>
>> Thanks
>>
>> Luigi
>>
>>
>> 2015-07-30 18:59 GMT+02:00 Eric24 <[email protected]>:
>>
>>> "Memory storage is for testing." Really? What if I don't want anything
>>> to persist? I would have assumed that memory storage in a cluster would be
>>> a good alternative (and fast) for something like what you might use
>>> memcached (or a variant) for?
>>> --Eric
>>>
>>>
>>> On Wednesday, July 29, 2015 at 8:30:15 AM UTC-5, Riccardo Tasso wrote:
>>>>
>>>> Memory storage is for testing.
>>>> If you want all your database in RAM use a plocal storage with HUGE
>>>> cache size.
>>>>
>>>> Cheers,
>>>>    Riccardo
>>>>
>>>> 2015-07-29 15:16 GMT+02:00 Arpit Jain <[email protected]>:
>>>>
>>>>> Hello Members,
>>>>>
>>>>> I have to perform a load test of orientDB write operation in memory
>>>>> storage distributed mode.
>>>>> i have started a cluster with 3 node.
>>>>> when i used  ./console.sh, it just shows my plocal databases
>>>>>
>>>>>                                orientdb {server=remote:localhost/}>
>>>>> list databases
>>>>>
>>>>>                                Found 2 databases:
>>>>>
>>>>>                                GratefulDeadConcerts (plocal)
>>>>>                                julyRecord (plocal)
>>>>>
>>>>> so my doubt is how to connect with my memory database and how to
>>>>> monitor it in distribute mode?
>>>>>
>>>>> Regards:
>>>>> Arpit
>>>>>
>>>>> --
>>>>>
>>>>> ---
>>>>> You received this message because you are subscribed to the Google
>>>>> Groups "OrientDB" group.
>>>>> To unsubscribe from this group and stop receiving emails from it, send
>>>>> an email to [email protected].
>>>>> For more options, visit https://groups.google.com/d/optout.
>>>>>
>>>>
>>>> --
>>>
>>> ---
>>> You received this message because you are subscribed to the Google
>>> Groups "OrientDB" group.
>>> To unsubscribe from this group and stop receiving emails from it, send
>>> an email to [email protected].
>>> For more options, visit https://groups.google.com/d/optout.
>>>
>>
>>

-- 

--- 
You received this message because you are subscribed to the Google Groups 
"OrientDB" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to [email protected].
For more options, visit https://groups.google.com/d/optout.

Reply via email to