Index-level TTL -- how do we track this feature request?

Hi --

I understand there is a feature request for index-level TTL. My question is, how do I track that feature request?


It'd be in here -

If you find it, please link it here so others can benefit :slight_smile: (I couldn't find it!)

To get the info, I opened a new feature in github. Then got this reply: "The existing feature request is at #2114. It was closed because curator can already do this and it's fairly easy to do so (delete + index selection). I'm going to close this issue as it seems like it's an exact duplicate of #2114. If the existing curator functionality doesn't fill your needs, feel free to +1 and/or comment in there and we can always consider reopening it." So, I added a "+1. Please! Yes, curator can do it, but it's way more efficient to set a ttl on the index at time of creation!". Anyone else can do the same here:

Hi Mark,

We developed a c# console app to manage our index retention until a TTL at the index level is an option. At a high level: The program accepts index type values and a date retention value. It searches through index names for any that match the pattern, then checks the date in the index name to see if it is older than the date retention value, and if so, deletes it. Has some safety nets built in, detailed logging, etc.

Some other folks in our company heard about it and asked for it because they had been looking for "exactly that", so I thought maybe others outside our company could benefit. Is this something that should be considered for addition to the elasticsearch github? If so, what's the process?


Curator does exactly that though, so I am not sure why you rebuilt it.

We tried to use curator for our purposes, but it wouldn't run as a scheduled task in a windows environment (running it manually worked), so we went another way. I tried to get help (Different results when running curator command at command prompt vs as a .bat file in windows) but it didn't go well.

Also, for the prefix option, does curator allow you to target multiple prefixes in the same run?

You could do this, but it would take a custom regular expression, via the --regex flag.

Do you know of people being able to have curator run in an automated fashion in windows environments? As mentioned in the thread mentioned above, we got different results running it manually vs as a .bat (couldn't get the .bat to work). I submitted details and logs but didn't get a reply.

I do not know. Is there a Windows equivalent of the cron service?

EDIT: It appears there are several possibilities: