Replication on data nodes

I was upgrading elasticsearch onmy cluster. Successfully been done, I face an issue,

the issue is on my data nodes when I see on the dashboard, one data node has all the original data while other has all the replicated data.
Below is the image:

the information of a single shard is below:

{
    "state": "STARTED",
    "primary": true,
    "node": "31mUy4upTjqVgE6mPZo9Zg",
    "relocating_node": null,
    "shard": 2,
    "index": "search_feedback",
    "allocation_id": 
    {
         "id": "VFyr5yfRTO692Lsytpi_nQ"
     }
}

If you see the screenshot the shards with the gray border are the original shards while the one's without are the replicas.

How can I make them reallocate such that a each data node has some original shards and some replicas, like the way they used to before I did the upgrade.

It does not really matter. Why do you think it does?

@dadoonet It just seems off since it's never happened that ways. Also I get an error while querying a particular index. with the error illegal_state_exception. below is a link

Also I get an error while querying a particular index. with the error illegal_state_exception. below is a link

This is not related.

It just seems off since it's never happened that ways.

You can bring up a new node, wait for a new distribution of the shards, then switch off this node and wait for the new distribution and you will probably end up with replicas and primaries mixed together. But I'd not care about this unless you really have performance problems or something like that.

Great! glad to know that, I was wondering if I have to reindex my data

Would you please be able to check out the issue that I've linked in my previous reply? It'll be of great help

I looked at it but I don't have the answer so I did not answer as I prefer to leave it non answered so you get a better chance to have someone else to look at it.

This topic was automatically closed 28 days after the last reply. New replies are no longer allowed.