Hi Michael,
with default cfg it's supposed to work, because you've:
"failureAvailableNodesLessQuorum": false,
So also with 1 node the cluster should continue working. I reproduced your
steps and works. How did you inserted record? console? Studio? App?
Lvc@
On 2 June 2014 05:18, Michael Campbell <[email protected]> wrote:
> Hate to bother but wondered if this was expected behavior or not.
>
> Using orientdb-community-1.7-20140522.195036-229-distribution.zip.
>
> Also using the default distributed config, but here it is. As I
> mentioned, I get the same effect with writeQuorum set to 2, or 1.
>
> {
> "autoDeploy": true,
> "hotAlignment": false,
> "offlineMsgQueueSize" : 0,
> "readQuorum": 1,
> "writeQuorum": 1,
> "failureAvailableNodesLessQuorum": false,
> "readYourWrites": true,
> "clusters": {
> "internal": {
> },
> "index": {
> },
> "*": {
> "servers" : [ "<NEW_NODE>" ]
> }
> }
> }
>
>
> On Fri, May 23, 2014 at 11:32 AM, Luca Garulli <[email protected]>
> wrote:
>
>> Hi,
>> what release are you using (Please say always the version you're using)?
>> Can you post here your distributed-config.json file?
>>
>> Lvc@
>>
>>
>>
>> On 23 May 2014 17:14, Michael Campbell <[email protected]>
>> wrote:
>>
>>> I'm playing with Orient in replicated mode. So far I have it working ok
>>> where I have 2 nodes on 2 boxes, and they see each other. When I make a
>>> change on one, the other sees it.
>>>
>>> I have writeQuorum to 2.
>>>
>>> So here's what I'm trying that doesn't meet my expectations, and I'm
>>> wondering if it's just my expectations that are wrong or something else.
>>> Since I have 2 nodes, I'll call them "one" and "two" for this example.
>>>
>>> 1. create a database on "one".
>>> 2. confirm the db exists on "two". (works)
>>> 3. create a class and property on "two"
>>> 4. confirm exists on "one" (works)
>>> 5. insert a row into class on "one"
>>> 6. confirm row exists on "two" (works)
>>> 7. Kill node "one"
>>> 8. insert a row in "two" (hangs...)
>>> 9. restart "one"
>>> 9.1 (insert from step 8 now completes)
>>> 10. see if row exists in "one" (it does not)
>>>
>>>
>>> I tried this with writeQuorum set to 2, and to 1.
>>>
>>> So my question is about this killing and restarting. Is this what it's
>>> supposed to be doing? What I'm trying to test is how to set up replication
>>> so that if one of the replicated stores fails, that when it comes back up
>>> it will catch up and get all the updates.
>>>
>>> Someone show me where I've gone wrong?
>>>
>>> --
>>>
>>> ---
>>> 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.
>
--
---
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.
