Commit graph

166 commits

Author SHA1 Message Date
martinsumner
4a8a2c1555 Code reduction refactor
An attempt to refactor out more complex code.

The Penciller clerk and Penciller have been re-shaped so that there
relationship is much simpler, and also to make sure that they shut down
much more neatly when the clerk is busy to avoid crashdumps in ct tests.

The CDB now has a binary_mode - so that we don't do binary_to_term twice
... although this may have made things slower ??!!?  Perhaps the
is_binary check now required on read is an overhead.  Perhaps it is some
other mystery.

There is now a more effiicient fetching of the size on pcl_load now as
well.
2016-10-08 22:15:48 +01:00
martinsumner
2055f8ed3f Add more complex snapshot test
This exposed another off-by-one error on startup.

This commit also includes an unsafe change to reply early from a rolling
CDB file (with lots of objects writing the hash table can take too
long).  This is bad, but will be resolved through a refactor of the
manifest writing:  essentially we deferred writing of the manifest
update which was an unnecessary performance optimisation.  If instead we
wait on this, the process is made substantially simpler, and it is safer
to perform the roll of the complete CDB journal asynchronously.  If the
manifest update takes too long, an append-only log may be used instead.
2016-10-07 10:04:48 +01:00
martinsumner
ad5aebe93e Further work on system tests
Another issue exposed with laziness in the using an incomplete ledger
when checking for presence during compaction.
2016-10-05 18:28:31 +01:00
martinsumner
d903f184fd Add initial end-to-end common tests
These tests highlighted some logical issues when scanning over databases
on startup, so fixes are wrapped in here.
2016-10-05 09:54:53 +01:00
martinsumner
507428bd0b Add initial system test
Add some initial system tests.  This highlighted issues:
- That files deleted by compaction would be left orphaned and not close,
and would not in fact delete (now deleted by closure only)
- There was an issue on stratup that the first few keys in each journal
would not be re-loaded into the ledger
2016-10-03 23:34:28 +01:00
martinsumner
c64d67d9fb Snapshot Work - Interim Commit
Some initial work to get snapshots going.

Changes required, as need to snapshot through the Bookie to ensure that
there is no race between extracting the Bookie's in-memory view and the
Penciller's view if a push_to_mem has occurred inbetween.

A lot still outstanding, especially around Inker snapshots, and handling
timeouts
2016-09-23 18:50:29 +01:00
martinsumner
a1c970a66a Manifest ordering
Be more explicit about manifest ordering to stop keys being laoded in
incorrect order
2016-09-19 15:56:35 +01:00
martinsumner
7c28ffbd96 Further bookie test - CDB optimisation and Inker manifest correction
Additional bookie test revealed that the persisting/reading of inker
manifests was inconsistent and buggy.

Also, the CDB files were inffeciently writing the top index table -
needed to be improved as this is blokicng on a roll
2016-09-19 15:31:26 +01:00
martinsumner
b452fbe27c End-to-end test
Changes to ensure working of first end-to-end test (with a single Key
and Value)
2016-09-15 18:38:23 +01:00
martinsumner
5127119669 First draft of bookie
Now with the most simple of tests
2016-09-15 15:14:49 +01:00
martinsumner
e73a5bbf31 WIP - First draft of Bookie code
First draft of untested bookie code
2016-09-15 10:53:24 +01:00
martinsumner
86666b1cb6 Scan over CDB file
Make scanning over a CDB file generic rather than specific to read-in of
active nursery log - open to be called as an external function to
support other scanning behaviour.
2016-09-09 15:58:19 +01:00
martinsumner
edfe9e3bed Improved testing
Improve testing of Penciller to show startup and shutdown with push,
merging and fetch
2016-09-08 14:21:30 +01:00
martinsumner
f0e1c1d7ea Basic GET/PUT and rolling in Inker
Add support to roll file on PUT in the inker
2016-09-06 17:17:31 +01:00
martinsumner
33f1efd576 Work on descriptions
Add further descriptions of roles following name changes.  Attempt to
simplify manifest management in the Penciller by assuming there is only
one Penciller's Clerk active - and so only one piece of work can be
ongoing
2016-08-02 17:51:43 +01:00
martinsumner
2bdb5fba6c Re-naming
Naming things is hard.  This change renames things based on the
Bookie/Inker/Penciller terminology
2016-08-02 13:44:48 +01:00