subversion-dev mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From Julian Foad <julianf...@apache.org>
Subject Re: FSFS replication and the rep cache
Date Tue, 21 Aug 2018 15:27:21 GMT
Daniel Shahaf wrote:
> Julian Foad wrote on Tue, 21 Aug 2018 14:12 +0100:
> >   * We could ignore the rep cache during commit-txn (in existing API: 
> > set fs_fs_data.rep_sharing_allowed = FALSE), then make a separate call 
> > to update the rep cache afterwards.
> 
> "A separate call to update the rep cache afterwards" --- something like
> `svnadmin backfill-rep-cache -r 40:50` --- could be generally useful, to
> admins who had disabled rep-cache or deleted rep-cache.db (intentionally
> or otherwise) and want to rebuild the rep-cache without a dump/load cycle.

Yes, I totally agree, that would be very useful functionality in its own right.

> (To be clear, I'm not opining that the logic should be in svnadmin; that
> was just a pseudocode example.)

Agreed. Also we might like to teach "svnadmin verify" to check whether the cache is complete.

> >   * We could change FSFS to allow selectively disabling part 2 (look up 
> > props) during the 'commit' step while keeping part 3 (update) enabled 
> > (split that flag into two), and disable part 2 only.
> > 
> > With these first two options, the rep cache would deduplicate only file 
> > contents, and that is fine. Deduplication of properties is relatively 
> > minor.
> > 
> >   * We could change FSFS such that commit-txn no longer depends on the 
> > rep-cache content, by moving the props deduplication to the txn-building 
> > phase.
> > 
> >   * We could ensure the rep cache is synchronized across repositories 
> > before each commit-txn.
> 
> Who's "We" in the last sentence?

WANdisco, possibly requiring additional support from Subversion. (I haven't thought about
how it might work.)

> > Could I please hear your thoughts? How appropriate might it be to make 
> > such changes in FSFS, if they are potentially beneficial for other users 
> > of the FSFS API, or other considerations?
> 
> Thanks for sharing the wider context.

-- 
- Julian

Mime
View raw message