M/S迁移数据到cluster报错

源节点是Master-slave架构的，版本是redis_version:3.0.5。
目标节点是cluster集群，版本是redis_version:3.2.7。
命令如下：

./redis-sync -m password@ip:port1-t ip2:port2 --db=0 --tmpfile-size=5gb

rdb sync完，等10+秒，就会抛panic了。。。

2018/10/25 15:43:23 libs.go:462: [PANIC] fetch redigo reply failed
[error]: MOVED 3693 10.19.xx.xx:7461
[stack]:
2 /data/go/src/github.com/CodisLabs/redis-port/cmd/libs.go:462
main.redigoGetResponse
1 /data/go/src/github.com/CodisLabs/redis-port/cmd/libs.go:588
main.doRestoreDBEntry.func3
0 /data/go/src/github.com/CodisLabs/redis-port/cmd/libs.go:314
main.(*Job).Run.func1
... ...
以为用的redis third_party源码包不对，之后换成了3.2的源码包，还是报错了。

还是说不支持这种异构实例的迁移？

Reproduce the reported migration from Redis 3.0.5 master-slave to Redis 3.2.7 cluster with the shown redis-sync command. Start by reading cmd/libs.go around redigoGetResponse at line 462 and doRestoreDBEntry at line 588, then inspect the third_party Redis client version used. Done means the supported behavior is established and the MOVED response no longer causes an unexplained panic.

Written by the indexing model from the issue text.