diff options
Diffstat (limited to 'doc/impl.html')
| -rw-r--r-- | doc/impl.html | 13 |
1 files changed, 1 insertions, 12 deletions
diff --git a/doc/impl.html b/doc/impl.html index b190d2c..dd09fea 100644 --- a/doc/impl.html +++ b/doc/impl.html @@ -57,15 +57,6 @@ These merges have the effect of gradually migrating new updates from the young level to the largest level using only bulk reads and writes (i.e., minimizing expensive seeks). -<h2>Large value files</h2> -<p> -Each large value (greater than 64KB by default) is placed in a large -value file (*.val) of its own. An entry is maintained in the log -and/or sorted tables that maps from the corresponding key to the -name of this large value file. The name of the large value file -is derived from a SHA1 hash of the value and its length so that -identical values share the same file. -<p> <h2>Manifest</h2> <p> A MANIFEST file lists the set of sorted tables that make up each @@ -220,9 +211,7 @@ So maybe even the sharding is not necessary on modern filesystems? compaction and at the end of recovery. It finds the names of all files in the database. It deletes all log files that are not the current log file. It deletes all table files that are not referenced -from some level and are not the output of an active compaction. It -deletes all large value files that are not referenced from any live -table or log file. +from some level and are not the output of an active compaction. </body> </html> |