Jake: https://hpccsystems.atlassian.net/browse/HPCC-23222

Wil do Jake, thank you!

- can you open a new JIRA(s) re. issues, as this issue can't be reopened.

But I think in summary, the UI presents 2 options "Replicate" and "Delayed replicate", but it's not clear why there are 2 options or what "Replicate" does.
And in reality the only type of replication spray has ever supported is delayed replication. I suspect when that option was added, it was supposed to replace "Replicate" - i.e. only clarify it's meaning.

The ability to replicate immediately, afaik, has never been available.
- can you confirm that?

Hi Jake,

I'm OK with this I guess, but when you say:

_"So as it stands, afaik - it is how it's always been (prior to ), files will be marked replicated regardless of the "Replicate:" option in Eclwatch."
_

So shouldn't we default that Replicate check box to ON ?

Thanks!

This JIRA/PR restored the behavior to how it was prior to HPC-13897.

That is, that despite what 'Replicate:' says, it should publish the file with the replicated flag on.
Which always meant - that there 'may' be a replicate copy, but the actual replication is done at some later stage when backupnode is run.

Assuming I'm right, and worth checking, regardless of what you choose when spraying "Is Replicated:" should say true for a sprayed logical file to a cluster that has replication.

As mentioned in the description of this JIRA:
> At a later time we can give users the option of spraying without any redundancy, with immediate redundancy or with delayed/async redundancy (See ).

So as it stands, afaik - it is how it's always been (prior to ), files will be marked replicated regardless of the "Replicate:" option in Eclwatch.
Getting spray to actually replicate immediately has never been possible afaik, and it's what was opened for.

- can you review the above summary, and - and check that it covers what you expect.