@@ -375,15 +375,25 @@ repl-diskless-sync no
375
375
# it entirely just set it to 0 seconds and the transfer will start ASAP.
376
376
repl-diskless-sync-delay 5
377
377
378
- # Replica can load the rdb it reads from the replication link directly from the
379
- # socket, or store the rdb to a file and read that file after it was completely
378
+ # -----------------------------------------------------------------------------
379
+ # WARNING: RDB diskless load is experimental. Since in this setup the replica
380
+ # does not immediately store an RDB on disk, it may cause data loss during
381
+ # failovers. RDB diskless load + Redis modules not handling I/O reads may also
382
+ # cause Redis to abort in case of I/O errors during the initial synchronization
383
+ # stage with the master. Use only if your do what you are doing.
384
+ # -----------------------------------------------------------------------------
385
+ #
386
+ # Replica can load the RDB it reads from the replication link directly from the
387
+ # socket, or store the RDB to a file and read that file after it was completely
380
388
# recived from the master.
389
+ #
381
390
# In many cases the disk is slower than the network, and storing and loading
382
- # the rdb file may increase replication time (and even increase the master's
391
+ # the RDB file may increase replication time (and even increase the master's
383
392
# Copy on Write memory and salve buffers).
384
- # However, parsing the rdb file directly from the socket may mean that we have
385
- # to flush the contents of the current database before the full rdb was received.
386
- # for this reason we have the following options:
393
+ # However, parsing the RDB file directly from the socket may mean that we have
394
+ # to flush the contents of the current database before the full rdb was
395
+ # received. For this reason we have the following options:
396
+ #
387
397
# "disabled" - Don't use diskless load (store the rdb file to the disk first)
388
398
# "on-empty-db" - Use diskless load only when it is completely safe.
389
399
# "swapdb" - Keep a copy of the current db contents in RAM while parsing
0 commit comments