[ 
https://issues.apache.org/jira/browse/SOLR-6273?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=14535654#comment-14535654
 ] 

Arcadius Ahouansou commented on SOLR-6273:
------------------------------------------

Note that this is just a question, not an objection.

The design blog talks only about 2 DCs being required.

Lets suppose that the DC1 and DC2 are both operational and updates being send 
to both of them.
If suddenly, the pipe between the two get broken for a couple of hours while 
updates still going into individual DC.
When the link is re-established, updates/replications from TransactionLog will 
fly from both DCs i.e DC1->DC2 and DC2->DC1



- How do we guaranty the order of execution of updates from TL? ... i.e when 
the link was broken, there have been deletion of doc#1 in DC1 followed by 
add/update of same doc#1 in DC2 

- In case DC1 lags far behind the other, full index replication (a la 
master-slave) may happen, meaning all updates done on DC1 will be overwritten 
by data from DC2 leading to data loss?

- Would a 3rd DC help make the system more redundant?


> Cross Data Center Replication
> -----------------------------
>
>                 Key: SOLR-6273
>                 URL: https://issues.apache.org/jira/browse/SOLR-6273
>             Project: Solr
>          Issue Type: New Feature
>            Reporter: Yonik Seeley
>            Assignee: Erick Erickson
>         Attachments: SOLR-6273.patch, SOLR-6273.patch, SOLR-6273.patch, 
> SOLR-6273.patch
>
>
> This is the master issue for Cross Data Center Replication (CDCR)
> described at a high level here: 
> http://heliosearch.org/solr-cross-data-center-replication/



--
This message was sent by Atlassian JIRA
(v6.3.4#6332)

---------------------------------------------------------------------
To unsubscribe, e-mail: dev-unsubscr...@lucene.apache.org
For additional commands, e-mail: dev-h...@lucene.apache.org

Reply via email to