mirror of
https://github.com/fluencelabs/redis
synced 2025-04-01 23:31:03 +00:00
Cluster: CLUSTER SETSLOT implementation comment updated.
Update the comment since the implementation details changed.
This commit is contained in:
parent
0064b1a583
commit
eb746ec408
@ -3320,9 +3320,11 @@ void clusterCommand(redisClient *c) {
|
|||||||
* to a new epoch so that the new version can be propagated
|
* to a new epoch so that the new version can be propagated
|
||||||
* by the cluster.
|
* by the cluster.
|
||||||
*
|
*
|
||||||
* FIXME: the new version should be agreed otherwise a race
|
* Note that if this ever results in a collision with another
|
||||||
* is possible if while a manual resharding is in progress
|
* node getting the same configEpoch, for example because a
|
||||||
* the master is failed over by a slave. */
|
* failover happens at the same time we close the slot, the
|
||||||
|
* configEpoch collision resolution will fix it assigning
|
||||||
|
* a different epoch to each node. */
|
||||||
uint64_t maxEpoch = clusterGetMaxEpoch();
|
uint64_t maxEpoch = clusterGetMaxEpoch();
|
||||||
|
|
||||||
if (myself->configEpoch == 0 ||
|
if (myself->configEpoch == 0 ||
|
||||||
|
Loading…
x
Reference in New Issue
Block a user