Pegasus crashed when I use the incorrect cluster

## Bug Report

Please answer these questions before submitting your issue. Thanks!

1. What did you do?
If possible, provide a recipe for reproducing the error.
```
./run.sh shell
>>> cc 23232323
```
2. What did you expect to see?
some Error messages,like 'Cluster dosen't exist'
3. What did you see instead?
Pegasus crashed
```
>>> cc 23232323
F2019-11-01 15:36:15.609 (1572593775609414389 4057) mimic.io-thrd.16471: replication_common.cpp:573:load_meta_servers(): assertion expression: servers.size() > 0
F2019-11-01 15:36:15.609 (1572593775609467709 4057) mimic.io-thrd.16471: replication_common.cpp:573:load_meta_servers(): no meta server specified in config [pegasus.clusters].23232323
./run.sh: 行 1621: 16471 已放弃 (核心已转储) ./pegasus_shell ${CONFIG} $CLUSTER_NAME
```

4. What version of Pegasus are you using?
Pegasus Shell 1.12.SNAPSHOT (404079781f91cd869949eb00c525155b330c9d7a) Release

Start by reproducing the invalid-cluster command with ./run.sh shell and cluster name 23232323, then inspect replication_common.cpp:573 and the Pegasus shell path shown in the report. Trace how a missing cluster configuration is handled. Done means the command reports that the cluster does not exist instead of crashing.

Written by the indexing model from the issue text.