David Turner d01436de3c

Copy checkpoint atomically when rolling generation (#35407) ...

Today when rolling a transog generation we copy the checkpoint from
`translog.ckp` to `translog-nnnn.ckp` using a simple `Files.copy()` followed by
appropriate `fsync()` calls. The copy operation is not atomic, so if we crash
at the wrong moment we can leave an incomplete checkpoint file on disk. In
practice the checkpoint is so small that it's either empty or fully written.
However, we do not correctly handle the case where it's empty when the node
restarts.

In contrast, in `recoverFromFiles()` we _do_ copy the checkpoint atomically.
This commit extracts the atomic copy operation from `recoverFromFiles()` and
re-uses it in `rollGeneration()`.

2018-11-23 08:43:34 +00:00

..

licenses

Upgrade to lucene-8.0.0-snapshot-67cdd21996 (#35816)

2018-11-22 15:42:59 +01:00

src

Copy checkpoint atomically when rolling generation (#35407)

2018-11-23 08:43:34 +00:00

build.gradle

Build: Fix official plugins list (#35661)

2018-11-19 10:17:05 -05:00