Hello,
I am having some doubts on what would be the correct configuration in a DataMiner cluster that has two elastic clusters geographically distributed.
In this specific case, we have a Failover pair with a CassandraCluster (6 nodes - 3 nodes in each datacenter) and 2 elastic clusters (3 nodes in each one).
Each elastic cluster is linked to a dataminer agent, as it is showed in the image below:
In this case, we want that one cluster is the replica of the other. That way, if one of the regions has a power outage, for example, the other one can take over.
Currently this is configured on db.xml as:
Zone 1 (Main agent):
<DataBases xmlns:xsi="http://www.w3.org/2001/XMLSchema-instance" xmlns:xsd="http://www.w3.org/2001/XMLSchema" xmlns="http://www.skyline.be/config/db">
<DataBase active="true" local="true" type="CassandraCluster" consistencyLevel="Quorum">
<DBServer>IPsOfCluster(ip1, ip2, ip3, ip4, ip5, ip6)</DBServer>
<DB>sldmadb</DB>
<UID>XXXX</UID>
<PWD>XXXX</PWD>
</DataBase>
<DataBase active="true" search="true" type="Elasticsearch">
<DBServer>IPs-zone1(ip7, ip8, ip9)</DBServer>
<UID>XXXX</UID>
<PWD/>
</DataBase>
</DataBases>
Zone 2 (Failover agent):
<DataBases xmlns:xsi="http://www.w3.org/2001/XMLSchema-instance" xmlns:xsd="http://www.w3.org/2001/XMLSchema" xmlns="http://www.skyline.be/config/db">
<DataBase active="true" local="true" type="CassandraCluster" consistencyLevel="Quorum">
<DBServer>IPsOfCluster(ip1, ip2, ip3, ip4, ip5, ip6)</DBServer>
<DB>sldmadb</DB>
<UID>XXXX</UID>
<PWD>XXXX</PWD>
</DataBase>
<DataBase active="true" search="true" type="Elasticsearch">
<DBServer>IPs-zone2(ip10, ip11, ip12)</DBServer>
<UID>XXXX</UID>
<PWD/>
</DataBase>
</DataBases>
Please take into account that the ips in this question are represented as ip1,2,3,…, but they are only placeholders, as in the actual configuration we have a normal IP 🙂
In this case, they are not using any DBConfiguration files. However, in the documentation, I can see that the supported configuration is using the DBConfiguration.xml file to use two different elastic clusters: Configuring multiple Elasticsearch clusters | DataMiner Docs
That way, can you please advise on what would be the best configuration for elastic in this case? It is expected to have any issue with the current configuration? How do the elastic clusters communicate with each other?
Thanks in advance for any help!
In the db.xml file we indeed no longer include the ElasticSearch db from 10.2.0 and 10.1.3 onwards.
Below is an example of a similar setup.
The data is offloaded to both Elastic clusters. As far as I'm aware they are not interlinked.
Hi Carolina, I’m not sure if there’s backward compatibility with the old way of configuring this, but to be on the safe side I’d suggest using the new method.
Ok, I’ll change it for the DBConfiguration.xml. Thanks!
Hello Jeroen, thanks for your reply!
This means that the current configuration will not work as we want to, right?