lucene-dev mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From "Jason Rutherglen (JIRA)" <>
Subject [jira] Commented: (LUCENE-2324) Per thread DocumentsWriters that write their own private segments
Date Wed, 05 Jan 2011 00:50:49 GMT


Jason Rutherglen commented on LUCENE-2324:

Taking a step back, I'm not sure flush control should be global, as flushing is
entirely per thread now? If we're adding a delete term for every DWPT, if one
is flushing do we wait or do we simply queue it up? I don't think we can wait
in the delete call for a DWPT to completely flush?

So we'll likely need to delete in a PerThreadTask that's executed on each
existing DWPT. How we guarantee concurrency seems a little odd here as what if
a new DWPT is spun up while we're deleting in another thread? Perhaps we should
simply spin up on DW init, the max thread state number of DWPTs? This way we
always have > 0 available, and we can perhaps lock on seq id when adding
deletes to all DWPTs (it's a fast call). We may want to simply skip any
flushing DWPTs when adding deletes?

In browsing the code, FlushControl isn't used in very many places. This'll get
a little bit more fleshed out when we integrate LUCENE-2573.

> Per thread DocumentsWriters that write their own private segments
> -----------------------------------------------------------------
>                 Key: LUCENE-2324
>                 URL:
>             Project: Lucene - Java
>          Issue Type: Improvement
>          Components: Index
>            Reporter: Michael Busch
>            Assignee: Michael Busch
>            Priority: Minor
>             Fix For: Realtime Branch
>         Attachments: LUCENE-2324-SMALL.patch, LUCENE-2324-SMALL.patch, lucene-2324.patch,
lucene-2324.patch, LUCENE-2324.patch, test.out
> See LUCENE-2293 for motivation and more details.
> I'm copying here Mike's summary he posted on 2293:
> Change the approach for how we buffer in RAM to a more isolated
> approach, whereby IW has N fully independent RAM segments
> in-process and when a doc needs to be indexed it's added to one of
> them. Each segment would also write its own doc stores and
> "normal" segment merging (not the inefficient merge we now do on
> flush) would merge them. This should be a good simplification in
> the chain (eg maybe we can remove the *PerThread classes). The
> segments can flush independently, letting us make much better
> concurrent use of IO & CPU.

This message is automatically generated by JIRA.
You can reply to this email to add a comment to the issue online.

To unsubscribe, e-mail:
For additional commands, e-mail:

View raw message