[Vertica报错]Spread error; can't determine which DBs are running; attempting to continue... '

以dbadmin用户启动vertica数据库时，报下面的错误：
Spread error; can't determine which DBs are running; attempting to continue... '
*** Starting database: noas ***
Enter password for [email protected] (2 attempts left):
    Spread does not seem to be running on 10.41.20.113.  The database will not be started on this host.

The following host(s) are not available:  10.41.20.113.
You should get them running first.  Operation can not be completed.
Database start up failed.  K-safe parameters not met.
Press RETURN to continue

查看vertica启动时的后台日志：/opt/vertica/log/adminTools-dbadmin.log
Nov  4 23:06:50  [adminTools] Use unique ports = False
Nov  4 23:06:50  [adminExec.__init__] Unique Ports: False
Nov  4 23:06:52  [spreadAdmin.gRDBFS]: EOF (spread not running?)         --发现spread没有启动
Nov  4 23:06:52  [spreadAdmin.gRDBFS]: EOF output was: 'Spread library version is 4.1.0^M
SP_error: (-2) Could not connect. Is Spread running?^M
^M
Bye.^M
', '<class 'vertica.utils.pexpect.EOF'>'
Nov  4 23:06:55  [adminCtrl.removeRunningDbs]: items '[('noas', 'noas', 0)]' runningDBs '[]'
Nov  4 23:06:55  [adminCtrl.dbStart]: candidates to start (no active) '[('noas', 'noas', 0)]'
Nov  4 23:06:56  [vsql.Passwords.Instance().getpass ] get password '[]'
Nov  4 23:06:58  [adminExec.getPortNo] no port assignment for noas, using 5433
Nov  4 23:06:58         Participating hosts:
Nov  4 23:06:58                 10.41.20.113
Nov  4 23:06:58                 10.41.20.116
Nov  4 23:06:58  [SSH.login user=dbadmin host=10.41.20.116]
Nov  4 23:07:01  [email protected]: ['cat /home/dbadmin/noas/v_noas_node0001_catalog/Epoch.log | head -3']
Nov  4 23:07:01  [email protected]: -- Executing -- cat /home/dbadmin/noas/v_noas_node0002_catalog/Epoch.log | head -3
Nov  4 23:07:01  [email protected]: cat /home/dbadmin/noas/v_noas_node0002_catalog/Epoch.log | head -3
Nov  4 23:07:01  [getLastGoodEpoch] trying to parse: Last good epoch: 0x8bf ended at '2013-11-14 11:54:22.087909+08'
Last good catalog version: 0x1419
K-safety: 0
Nov  4 23:07:01  KSafety: 0 Total hosts:2 Down hosts:[]
Nov  4 23:07:01  [email protected]: /etc/init.d/spreadd status
Nov  4 23:07:01  [email protected]: ['0', ['4938', '\r\x1b[80C\x1b[10D\x1b[1;32mrunning\x1b[m\x0f']]
Nov  4 23:07:01         Spread does not seem to be running on 10.41.20.113.  The database will not be started on this host.
You should get them running first.  Operation can not be completed.
Nov  4 23:07:01  Database start up failed.  K-safe parameters not met.
Nov  4 23:09:42  [SSH.logout]
Nov  4 23:07:01  [email protected]: ['0', ["Last good epoch: 0x8bf ended at '2013-11-14 11:54:22.087909+08'", 'Last good catalog version: 0x1419', 'K-safety: 0']]
Nov  4 23:07:01  [getLastGoodEpoch] trying to parse: Last good epoch: 0x8bf ended at '2013-11-14 11:54:22.087909+08'
Last good catalog version: 0x1419
K-safety: 0
Nov  4 23:07:01  [getLastGoodEpoch] found epoch '2239' at timestamp '2013-11-14 11:54:22.087909+08' with catalog version '5145' and ksafety '0'
Nov  4 23:07:01  [getLastGoodEpoch] trying to parse: Last good epoch: 0x8bf ended at '2013-11-14 11:54:22.087909+08'
Last good catalog version: 0x1419
K-safety: 0
Nov  4 23:07:01  [getLastGoodEpoch] found epoch '2239' at timestamp '2013-11-14 11:54:22.087909+08' with catalog version '5145' and ksafety '0'
Nov  4 23:07:01  KSafety: 0 Total hosts:2 Down hosts:[]
Nov  4 23:07:01  [email protected]: /etc/init.d/spreadd status
Nov  4 23:07:01  [email protected]: ['0', ['4938', '\r\x1b[80C\x1b[10D\x1b[1;32mrunning\x1b[m\x0f']]
Nov  4 23:07:01         Spread does not seem to be running on 10.41.20.113.  The database will not be started on this host.
Nov  4 23:07:01
The following host(s) are not available:  10.41.20.113.
You should get them running first.  Operation can not be completed.
Nov  4 23:07:01  Database start up failed.  K-safe parameters not met.
Nov  4 23:09:42  [SSH.logout]

经过分析，发现vertica集群中另外一台机器没有启动spreadd进程。

linux113:/opt # /etc/init.d/spreadd status
                                                                                                                         dead
If having trouble starting spread, try using 'spreadd stop' to clear state,
check /opt/vertica/config/vspread.conf and spread logs in /tmp/spread_*
and /var/log/spreadd.log, also check that /var has not run out of space.

解决办法：
启动spreadd进程：
linux113:/opt # /etc/init.d/spreadd start