Question

我试图从完整的AMI图像中将不同的EC2主机（使用不同的ip addr）上的MemSQL，Community Edition，单主机群集调出来。我已尝试How to change IP addresses of memsql nodes中的建议，但我无法使其发挥作用。我可以通过Web界面连接到MemSQL-ops。它在新的本地主机上显示具有新IP地址的MASTER AGGREGATOR。但是，它仍在寻找旧IP地址的群集。我试图取消监控/监控MASTER AGGREGATOR和LEAF。命令和错误的输出如下：

10.0.0.232: ~$ sudo memsql-ops memsql-start --all
Starting cluster
MemSQL node BD090F1 failed to start: Failed to connect to MemSQL node BD090F1: No error in tracelog
10.0.0.232: ~$ sudo memsql-ops memsql-unmonitor
 Index  ID       Agent Id  Process State  Cluster State  Role    Host        Port  Version
 1      6A15216  Aec14b6   NOT RUNNING    CONNECTED      MASTER  10.0.0.181  3306  5.5.11
 2      BD090F1  Aec14b6   NOT RUNNING    CONNECTED      LEAF    10.0.0.181  3307  5.5.11
Select an option: 2
Would you additionally like to remove the leaf node from the MemSQL cluster? (You can specify this behavior with --remove-leaf.) [y/n] n
2017-01-18 15:38:29: J5ba705 [INFO] Stopping monitoring for MemSQL node BD090F1
2017-01-18 15:38:29: J5ba705 [INFO] No longer monitoring MemSQL node BD090F1
10.0.0.232: ~$ sudo memsql-ops memsql-monitor -h 127.0.0.1 -P 3307
2017-01-18 15:38:53: Jb697d2 [INFO] Monitoring MemSQL node at 10.0.0.232:3307
2017-01-18 15:38:55: Jb697d2 [ERROR] Failed to connect to MemSQL at 10.0.0.232:3307:
10.0.0.232: ~$ sudo memsql-ops memsql-monitor -h 127.0.0.1 -P 3306
2017-01-18 15:40:32: J18a2de [INFO] Monitoring MemSQL node at 10.0.0.232:3306
2017-01-18 15:40:32: J18a2de [ERROR] Failed to connect to MemSQL at 10.0.0.232:3306:
10.0.0.232: ~$ sudo memsql-ops memsql-start --all
Starting cluster
MemSQL node 6A15216 failed to start: Failed to connect to MemSQL node 6A15216: No error in tracelog

Answer 1

看起来memsql ops正在使用memsql-ops代理的ip地址。要修复它，您需要一个额外的步骤：

memsql-ops memsql-unmonitor for each memsql node
sudo memsql-ops stop
sudo memsql-ops start --host 127.0.0.1
memsql-ops memsql-monitor [-h 127.0.0.1] [-P PORT]

无法使用AMI将MemSQL Community Edition移动到新的EC2主机

1 个答案: