Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Bogn: Enable value-log support with bubt. #72

Closed
prataprc opened this issue Mar 21, 2018 · 0 comments
Closed

Bogn: Enable value-log support with bubt. #72

prataprc opened this issue Mar 21, 2018 · 0 comments

Comments

@prataprc
Copy link
Member

Once value log support is available in Bubt implementation, implement that feature in bogn as well.

@prataprc prataprc changed the title Bogn: Enalbe value-log support with bubt. Bogn: Enable value-log support with bubt. Apr 4, 2018
prataprc added a commit that referenced this issue Apr 4, 2018
If value log is enabled (by supplying vblocksize > 0), we
pre-emptively co-locate value-log files along with z-index files.

Application can use AppendValuelogs() method to supply value-log
files to append values onto existing files, instead of creating
new files. `vblocksize` should be same as used for supplied valuelogs.
prataprc added a commit that referenced this issue Apr 4, 2018
prataprc added a commit that referenced this issue Apr 5, 2018
Value log support #72.

* Use payload instead of footprint to decided whether or not to warmup
  from disk during boot.
* Log payload instead of footprint.

Flushing to disk and compaction of disk files need to be fine tuned and
prioritized, make the code better maintained #73.

* For flushing - "fresh", "aggressive", "fallback", "merge" are the order
  of priority
* For compaction - "none", "aggressive", "ratio", "period", "self" are
  the order of priority.

Cleanup validatedisklevel(), bubt.Snapshot.Validate() does most of the
job.
prataprc added a commit that referenced this issue Apr 14, 2018
Similar to api.Iterator for full table scan, api.EntryIterator is
created. EntryIterator will return an encapsulated object for each
entry in the index. The encapsulated object shall implement IndexEntry
interface. The basic idea behind IndexEntry is to allow lazy loading of
value bytes, which can save read amplification when two iterators
are to be merge sorted.

Implement ScanEntries, EntryIterator, IndexEntry for llrb.LLRb,
llrb.MVCC, bubt.Snapshot.

And few other developmental fixes.
prataprc added a commit that referenced this issue Apr 15, 2018
prataprc added a commit that referenced this issue Apr 15, 2018
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
None yet
Projects
None yet
Development

No branches or pull requests

1 participant