zookeeper-user mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From Alexander Shraer <shra...@gmail.com>
Subject Re: Incrementally bootstrapping a 3.5.0-alpha cluster?
Date Sat, 27 Jun 2015 03:30:35 GMT
Hi Raul,

Not sure - it seems that we have too few logging messages as it is since
often we can't figure out what happened even when we have the logs :) Maybe
the message should just clarify that its normal - server 1 contacted server
2 but since it has smaller id we expect this connection to be closed and
server 2 should establish connection soon. Not sure how to say this
concisely...

Alex

On Fri, Jun 26, 2015 at 5:32 PM, Raúl Gutiérrez Segalés <rgs@itevenworks.net
> wrote:

> Hi Alex,
>
> On 25 June 2015 at 14:36, Alexander Shraer <shralex@gmail.com> wrote:
>
> > This message itself doesn't indicate a failure, its quite normal. But if
> > you have a situation where the ensemble gets stuck or doesn't elect a
> > leader, please open a jira and post your server logs.
> >
>
> Do you think we should make:
>
>             LOG.info("Have smaller server identifier, so dropping the " +
>                      "connection: (" + sid + ", " + self.getId() + ")");
>
> a LOG.debug  statement instead? I see how people can see that as some sort
> of error condition...
>
>
> -rgs
>

Mime
  • Unnamed multipart/alternative (inline, None, 0 bytes)
View raw message