incubator-blur-user mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From Bradley Johnson <bjohn...@volumeintegration.com>
Subject Re: blur command to see if it is working
Date Fri, 13 May 2016 17:03:12 GMT
I know what is happening by looking at the logs, I just don't want to have
to look at the logs.

I just want an easy way to trigger alerts in an alerting agent if shards
start dying off or being marked as bad connections.

I will try issuing queries to each shard server to see if that will work
for us.


On Fri, May 13, 2016 at 12:34 PM, Aaron McCurry <amccurry@gmail.com> wrote:

> If everything is up and running and there is no activity then you should
> check the shard logs to see if there is an issue on a particular node.  One
> thing that you could try is to connect to a single shard server via the
> shell
>
> >blur shell shard-server:40020
>
> Then try to execute a query from there, if it returns then likely that
> server is fine.  However there isn't an easy way to walk through all the
> shards and tell if one or more are not responding.  Without seeing some
> logs it's hard to say what's happening.
>
> Aaron
>
>
> On Fri, May 13, 2016 at 9:49 AM, Bradley Johnson <
> bjohnson@volumeintegration.com> wrote:
>
> > Could you tell me a command I could run to ask our blur cluster if it is
> up
> > and working?
> >
> > At the moment the controllers have lost contact with all the shard
> servers.
> > BlurClientManager is reporting all connections to all shards are bad in
> the
> > logs. When I go to the blur console it still is displaying that all
> > Zookeepers, Controllers and Shards are up.
> >
> > Is there a blur command to check to see if blur is actually working?
> >
> > The commands of
> > blur shell
> >
> > top - shows all the shards with 0 activity
> > controllers - shows all the controllers
> > shards - shows all the shards
> > list - lists the tables as enabled
> >
> > but in reality nothing is working.
> >
> > The html consol gui shows everything is up accept for the tables. Lots of
> > log errors show up under the log button. unknown error during remote call
> > to node [xxxxxxxxx]
> >
> > --
> > Brad Johnson
> > (571) 248-1566
> > Volume Integration <http://volumeintegration.com>
> > Volume Labs <http://volumelabs.net>
> >
>



-- 
Brad Johnson
(571) 248-1566
Volume Integration <http://volumeintegration.com>
Volume Labs <http://volumelabs.net>

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