如何恢复快照(rdb 文件)从另一台机器复制的重写数据？

There is nothing specific to do. Just install the redis server on the new machine, and edit the configuration file. You just need to change the following parameters to point to the location of the dump file you have just copied.

# The filename where to dump the DB
dbfilename mydump.rdb


# The working directory.
#
# The DB will be written inside this directory, with the filename specified
# above using the 'dbfilename' configuration directive.
#
# Also the Append Only File will be created inside this directory.
#
# Note that you must specify a directory here, not a file name.
dir /data/mydirectory/

Finally, the redis server can be started in the normal way.

start redis on your second server, like so:

$ > redis-server /path/to/my/redis/configuration/file/redis.conf

when redis starts, it will find your rdb file because it will look for the name and file path in the configuration file (redis.conf) that you supply when start the redis server, as above.

to supply the file name and path, just edit two lines in the redis.conf file template (supplied in the root directory of the redis source. Save your revised version as redis.conf in the directory location that you supplied upon starting the server.

You will find the settings you need in the redis.conf template in the source top-level directory, at lines 127 and 137 (redis version 2.6.9).

# The filename where to dump the DB
dbfilename dump.rdb


# The working directory
dir ./

as you can see, defaults are provided for both settings

so just change the first of these two lines (127) to identify your rdb file and in the second (137) substitute the default "./" for the actual file path for your snapshot rdb file; save the redis.conf with your changes, and start redis passing in this new conf file.

Or you can:

1. Stop your redis server / instance, eg., service redis6379 stop
2. Copy the dump.rdb file to the right location, eg., cp /path/to/dump-6379.rdb /var/lib/redis/dump-6379.rdb. Give it the right permissions (user:group should be redis:redis and mode 644)
3. Start your redis server / instance, eg., service redis6379 start

It is important that you stop the redis server before copying the file to the right location, because Redis saves a snapshot before terminating, so it will replace your file.

Besides, you might want to back up the existing dump.rdb file first.

For databases where the appendonly flag is set to no, you can do the following:

1. Stop Redis (because Redis overwrites the current rdb file when it exits).
2. Copy your backup rdb file to the Redis working directory (this is the dir option in your Redis config). Also, make sure your backup filename matches the dbfilename config option.
3. Start Redis.

If, on the other hand, you need to restore an rdb file to the appendonly database, you should do something along the lines of:

1. Stop Redis (because Redis overwrites the current rdb file when it exits).
2. Copy your backup rdb file to the Redis working directory (this is the dir option in your Redis config). Also, make sure your backup filename matches the dbfilename config option.
3. Change the Redis config appendonly flag to no (otherwise Redis will ignore your rdb file when it starts).
4. Start Redis.
5. Run redis-cli BGREWRITEAOF to create a new appendonly file.
6. Restore the Redis config appendonly flag to yes.

Specifically, this is the relevant bit of documentation from the Redis config file comments:

# Note that you can have both the async dumps and the append only file if you
# like (you have to comment the "save" statements above to disable the dumps).
# >> Still if appendonly mode is enabled Redis will load the data from the
# >> log file at startup ignoring the dump.rdb file.

try set appendonly no. In My case, *.aof file was empty(0 byte), must set appendonly=no then make it load the dump.rdb

Assuming that you run Redis 2.6 or higher, your Redis snapshot filename is dump.rdb, and it exists in the directory /home/user/dbs, the following command would do the trick:

redis-server --dbfilename dump.rdb --dir /home/user/dbs

Relevant section from the official documentation: Passing arguments via the command line

I would like to add here a tiny detail that did not get mentioned and I will not use config file but instead specify everything in the command line.

When both mydump.rdb and appendonly.aof files are specified when starting redis-server, it will be the appendonly.aof file that wins such that the data from appendonly.aof gets loaded. For example:

redis-server --dbfilename mydump001.rdb --dir /data --appendonly yes

The above start invocation will use the /dir location to find the presence of mydump001.rdb or appendonly.aof files. In this case, redis-server will load the contents from appendonly.aof. If appendonly.aof does not exists, it will create an empty /data/appendonly.aof and the redis-server will be empty.

If you want to load a specific dump file, you can do:

redis-server --dbfilename mydump001.rdb --dir /data

I added this answer coz it is not obvious which is which. In the presence of 2 backup files, and this is often not mentioned.

This solution work with redis-cluster, but should work too with redis.

Install this dependencie https://github.com/sripathikrishnan/redis-rdb-tools

pip install rdbtools python-lzf

after that execute this

rdb -c protocol /path/to/dump.rdb | redis-cli -h host -p port --pipe

If this is a cluster, the port should the master`s port.

Install https://github.com/leonchen83/redis-rdb-cli

rmt -s ./your-dump.rdb -m redis://host:port -r