How to go to offline mode

I want to reindex my index in offline mode in elasticsearch(6x..) inorder to avoid any errors or timeout issues. How do I do that?
I am new to online and offline terms.

I have no idea of what this means. Where did you see those terms?

@dadoonet Like offline mode in sql database is possible. Isn't there any ways of using ES without connecting to the server, then make changes like reindex, and get back online and update ?

"offline indexing generally refers to indexing in batch from a "source of truth" database, so if you had all of your records stored in redshift you might have a batch job to index all of the records from redshift. On-line indexing is indexing records/events "as they happen", so you might have location information or some other type of updates coming in on a stream (say kafka) and then you would want those updates to be reflected in your search index without having to wait to update all of the records in a batch"

I found the above information from one of the youtube comment section. Presenter has used those terms:

And following is from StackOverflow:

I'm just trying to understand why you'd need that.

But here are some advices that might be useful.

Use index aliases. So you can define an alias "foo" which goes to index "foo1".
You use this alias to index and search for documents. So instead of running a query like GET foo1/_search you run GET foo/_search.

Then you can create a new index name foo2. Index the documents in it.
Once done, you can switch the alias foo from foo1 to foo2.

The users will now searching in foo2 when doing GET foo/_search.
Then you can simply DELETE foo1.

I understood that alias can help make the search process better, but does it help on re-indexing documents? Can I use aliases to make re-indexing process faster? Right now I am trying to re-index my 5 TB document from old-index to new-index using reindex API in kibana. But, it always fails to complete since it takes hours and hours to re-index all the documents. From my 20million documents, max I have seen so far is 1.3 million document re-indexed(after 4-5 hours) and it stops. How do I solve this?

That's another question. Could you open a new question about the reindex slowness?

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