accumulo-notifications mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From "Christopher Tubbs (JIRA)" <>
Subject [jira] [Commented] (ACCUMULO-3918) Different locality groups can compact with different iterator stacks
Date Thu, 25 Jun 2015 17:33:04 GMT


Christopher Tubbs commented on ACCUMULO-3918:

How is this different than between two files? You can't know exactly when the ZK props changed
or when they propagated to the tserver. The fact that two locality groups share the same file
says nothing about how those locality groups are expected to behave with respect to each other.

I can understand the desire to have a consistent iterator stack for a single file... I just
can't see how it would affect the user experience, because they don't know which data gets
written to which file, or which locality groups are co-located in the same files anyway (we
don't expose that to them, and it's not formally defined, nor is it guaranteed to not change).
If this is considered a bug, then why isn't it considered a bug that this can happen between
sequential compactions to different files (rather than just different locality groups)?

I guess I'm just not understanding the use case.

FWIW (not much), I typically advise people in situations where this matters, to take the table
offline, change the table configuration, and then bring it back online.

> Different locality groups can compact with different iterator stacks
> --------------------------------------------------------------------
>                 Key: ACCUMULO-3918
>                 URL:
>             Project: Accumulo
>          Issue Type: Bug
>          Components: tserver
>    Affects Versions: 1.6.0
>            Reporter: John Vines
> While looking through the compactor code, I noticed that we load the iterator stack for
each locality group written and drop it when we're done. This means if a user reconfigures
iterators while a locality group is being written, the following locality groups will be compacted
inconsistently with the rest of the file.
> We should really read the stack once and be consistent for the entire file written.

This message was sent by Atlassian JIRA

View raw message