Fix reindex from remote clearing scroll #22525
Merged
Conversation
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Member
Author
|
@javanna, can you have a look at this one? It is another thing with reindex-from-remote and being super careful with the client. |
javanna
reviewed
Jan 10, 2017
Member
There was a problem hiding this comment.
what happened to all of these tests? They are already in the docs so we can remove them from here?
Member
Author
There was a problem hiding this comment.
Yeah. Sorry, I meant to leave that comment. I made those before we ran the docs snippets. I should have removed them in another PR but I figured I could leave a comment and defend it. I promptly forgot.
Member
There was a problem hiding this comment.
ok I assume that we got test coverage besides docs tests and those ones were copied from the docs just to make sure that the docs were correct, which is not needed anymore?
Reindex-from-remote had a race when it tried to clear the scroll. It first starts the request to clear the scroll and then submits a task to the generic threadpool to shutdown the client. These two things race and, in my experience, closing the scroll generally loses. That means that most of the time reindex-from-remote isn't clearing the scrolls that it uses. This isn't the end of the world because we flush old scroll contexts after a while but this isn't great. Noticed while experimenting with elastic#22514.
nik9000
force-pushed
the
reindex_from_remote_close_carefully
branch
from
0a60a04 to
b75edad
Compare
nik9000
added a commit
that referenced
this pull request
Jan 10, 2017
Reindex-from-remote had a race when it tried to clear the scroll. It first starts the request to clear the scroll and then submits a task to the generic threadpool to shutdown the client. These two things race and, in my experience, closing the scroll generally loses. That means that most of the time reindex-from-remote isn't clearing the scrolls that it uses. This isn't the end of the world because we flush old scroll contexts after a while but this isn't great. Noticed while experimenting with #22514.
nik9000
added a commit
that referenced
this pull request
Jan 10, 2017
Reindex-from-remote had a race when it tried to clear the scroll. It first starts the request to clear the scroll and then submits a task to the generic threadpool to shutdown the client. These two things race and, in my experience, closing the scroll generally loses. That means that most of the time reindex-from-remote isn't clearing the scrolls that it uses. This isn't the end of the world because we flush old scroll contexts after a while but this isn't great. Noticed while experimenting with #22514.
lcawl
added
:Distributed Indexing/CRUD
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
4 participants
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
Reindex-from-remote had a race when it tried to clear the scroll. It
first starts the request to clear the scroll and then submits a task
to the generic threadpool to shutdown the client. These two things
race and, in my experience, closing the scroll generally loses. That
means that most of the time reindex-from-remote isn't clearing the
scrolls that it uses. This isn't the end of the world because we
flush old scroll contexts after a while but this isn't great.
Noticed while experimenting with #22514.