You are viewing a plain text version of this content. The canonical link for it is here.
Posted to users@rocketmq.apache.org by 赵剑 <it...@gmail.com> on 2017/07/13 03:55:17 UTC
Master-slave synchronization big message write failure how to handle
Hello,
The current cluster version 4.1.0, 2m-2s-sync, the producer defaults
more than 4k for message compression; write more than 500k size message
client return FLUSH_SLAVE_TIMEOUT, the failure rate is very high, ask how
to optimize the cluster to solve this problem. Thank you.
Re: Master-slave synchronization big message write failure how to
handle
Posted by Zhanhui Li <li...@gmail.com>.
1. Please check value of syncFlushTimeout which defines how much time master waits for slave to replicate date before claiming syncing salve timeout. If you set this value too small, you may see send result as described.
2. Double check logs of master and slave brokers, make sure connection between them is stable. If the connection is unstable, brokers may spend quite a lot of time establishing new connections.
> 在 2017年7月13日，下午12:03，zuhouhuwj49846@touna.cn 写道：
>
> maxMessageSize=524288 #最大消息大小，默认512K
>
>
>
>
> From: 赵剑 <ma...@gmail.com>
> Date: 2017-07-13 11:55
> To: users <ma...@rocketmq.incubator.apache.org>
> Subject: Master-slave synchronization big message write failure how to handle
> Hello,
> The current cluster version 4.1.0, 2m-2s-sync, the producer defaults more than 4k for message compression; write more than 500k size message client return FLUSH_SLAVE_TIMEOUT, the failure rate is very high, ask how to optimize the cluster to solve this problem. Thank you.
Re: Master-slave synchronization big message write failure how to handle
Posted by "zuhouhuwj49846@touna.cn" <zu...@touna.cn>.
maxMessageSize=524288 #最大消息大小，默认512K
From: 赵剑
Date: 2017-07-13 11:55
To: users
Subject: Master-slave synchronization big message write failure how to handle
Hello,
The current cluster version 4.1.0, 2m-2s-sync, the producer defaults more than 4k for message compression; write more than 500k size message client return FLUSH_SLAVE_TIMEOUT, the failure rate is very high, ask how to optimize the cluster to solve this problem. Thank you.