[Gluster-users] Help with some socket related logwarnings

[Carl Boberg]

Ok,

The line   "option transport.socket.read-fail-log off" was not in my
/etc/glusterfs/glusterd.vol file on neither of the servers.
What does this line do exactly?
FYI. My install is just a basic rpm install made by following the docs. I
have not added anything to the config files neither by glusterd command or
by direct editing.

I have added the read-fail-log off option and restarted glusterd on both
servers.

I still get this in the cli.log about once every minute on one of the
servers:

[2012-02-27 10:48:55.332085] W [rpc-transport.c:606:rpc_transport_load]
0-rpc-transport: missing 'option transport-type'. defaulting to "socket"
[2012-02-27 10:48:55.338023] I
[cli-cmd-volume.c:1077:cli_check_gsync_present] 0-: geo-replication not
installed
[2012-02-27 10:48:55.338371] I
[cli-rpc-ops.c:251:gf_cli3_1_list_friends_cbk] 0-cli: Received resp to
list: 0
[2012-02-27 10:48:55.338450] I [input.c:46:cli_batch] 0-: Exiting with: 0

And this in the etc log at the same interval:

[2012-02-27 10:48:55.338244] I
[glusterd-handler.c:750:glusterd_handle_cli_list_friends] 0-glusterd:
Received cli list req
[2012-02-27 10:48:55.339195] W [socket.c:1494:__socket_proto_state_machine]
0-socket.management: reading from socket failed. Error (Transport endpoint
is not connected), peer (127.0.0.1:1023)

Cheers
---
Carl Boberg
Operations

Memnon Networks AB
Tegnérgatan 34, SE-113 59 Stockholm

Mobile: +46(0)70 467 27 12
www.memnonnetworks.com



On Mon, Feb 27, 2012 at 09:56, Amar Tumballi <amarts at redhat.com> wrote:

> On 02/26/2012 01:30 AM, Brian Candler wrote:
>
>> On Sat, Feb 25, 2012 at 08:27:06PM +0100, Carl Boberg wrote:
>>
>>>    [socket.c:1494:__socket_proto_**state_machine] 0-socket.management:
>>>    reading from socket failed. Error (Transport endpoint is not
>>>    connected), peer ([2]127.0.0.1:1020)
>>>    And I cant figure out why I get the socket error with 127.0.0.1
>>>    I have checked all network configurations/dns on the server.
>>>
>>
>> Including /etc/hosts? That is, you don't have an entry like
>>
>> 127.0.0.1  myhostname
>>
>> (Some distros have a habit of adding an entry like that for you)
>> ______________________________**______
>>
>
> Can you please check if your 'glusterd.vol' file (located in
> /etc/glusterfs/glusterd.vol) has below lines?
>
> "    option transport.socket.read-fail-log off"
>
> If not, you can go ahead and add it.
>
> Dan, to answer your question, this log in 'glusterd.log' file is not
> serious. But it gives a hint of possible network issues in Filesystem logs
> (like brick logs and nfs logs and native client logs), as in Filesystem
> setup we need to be connected all the time.
>
>
> Regards,
> Amar
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://supercolony.gluster.org/pipermail/gluster-users/attachments/20120227/8e54ef2d/attachment.html>