copy ru
To instantiate or repair a follower member of a replication unit you can copy another follower in that replication unit from another shard.
Syntax
COPY [RU|REPLICATION_UNIT] -RU ru_id
[-SOURCE source_shard_name] -TARGET target_shard_name
[-REPLACE repl_shard_name]
[-timeout timeout]
[-force]
Options
Table 2-37 GDSCTL copy ru Options
| Option | Description |
|---|---|
-force |
Allow the operation to bypass RAFT replication role checks. |
-replace |
Removes the replication unit follower
specified by |
-ru |
Replication unit ID |
-source |
Name of the source shard. |
-target |
Name of the target shard . |
-timeout |
Timeout of connection retention between the time FAN event is sent to clients and chunk going read-only/down (seconds). |
-verbose |
Enable verbose mode. |
Usage Notes
The source shard and target shard shouldn't be the replica leader for the given replication unit. If a member of the specified replication unit exists on the target shard, it is replaced by full replica of replication unit from the source shard.
If -source is not specified, then an existing follower of the
replication unit will be chosen automatically as the source shard.
If the target database doesn't already contain a member of the specified
replication unit, then the total number of replicas for the given replication
unit should be below replication factor, unless -replace is
specified.
If -replace is specified, the replication unit member is removed
from the specified database.
Examples
GDSCTL> copy ru -ru 1 -source sh1 -target sh2