Seelog queue overflow: more than 10000 messages in the queue. Flushing.

[carl-leopard]

短时间内产生大量的日志导致服务器崩溃

[1522140840]

这个问题解决了吗？我昨天遇到了【cihub/seelog是新代码】

[carl-leopard]

没有， 短时间产生大量的日志， 会导致日志丢失

[huaapp]

https://github.com/cihub/seelog/wiki/Logger-types

[huaapp]

We've just recommended the rule for async timer's interval tuning: the average interval of message consuming should be less than the average interval of message producing.

Current log messages limit in the queue is 10000.

But how do you know whether you set it correct and unwanted flushes don't happen? This is pretty simple: Seelog will write a following message to the stdout (Seelog uses stdout to output its internal errors, messages, etc.): "Seelog queue overflow: more than 10000 messages in the queue." So if you are tuning your async timer interval and want to see whether it is all okay with the queue check the flush messages from Seelog in your stdout.

[carl-leopard]

thanks.