[Gluster-users] Gluster 3.3.0 doesn't see neighbour

[Vladimir Yakovlev]

Hi Everyone,

Problem has been SOLVED.

My colleague was playing with hearbeat on this machine and didn't forget to
reset net.ipv4.ip_nonlocal_bind = 0. I fixed it, did sysctl -p and now:
# gluster peer probe 192.168.1.193
Probe successful

Thanks, guys, for great support and sorry for keep you busy ;-)

BR,
Vladimir.

On Mon, Jul 2, 2012 at 6:34 PM, Vladimir Yakovlev <mrquesty at gmail.com>wrote:

> Hi Philip,
>
> I tried IPs instead of hostnames at the beginning, but it didn't help:
>
> [root at host1 ~]# gluster peer probe 192.168.1.193
> Probe on localhost not needed
>
> My assumption is smth wrong smwhere else. Donno where - no ideas at all;
> the worst is i have a set of test machines, where everything works
> perfectly... :(
>
> There are two switches between these machines, but they allow Gluster
> traffic responsively.
>
> Hope, we can struggle this together ;-)
>
> Thanks, guys, for quick responses.
>
> Regards,
> vy
>
> On Mon, Jul 2, 2012 at 6:27 PM, Philip Poten <philip.poten at gmail.com>wrote:
>
>> Hi Vladimir
>>
>> To circumvent any remaining /etc/hosts abnormalities, try with IPs
>> only. if that doesn't work either, you have a problem outside name
>> resolution. If it works with IPs, the problem is the resolver.
>>
>> Note that nscd caches answers, also the "host" command uses DNS
>> responses, /not/ the /etc/hosts (nsswitch.conf order resp.) file - but
>> gluster does, as does "ping" and anything else using gethostbyname(3)
>>
>> hth,
>> Philip
>>
>> 2012/7/2 Vladimir Yakovlev <mrquesty at gmail.com>:
>> > Thanks, Brian, for the hint.
>> > I've changed /etc/hosts with respect to your comment, but it didn't help
>> > either.
>> >
>> > The problem (from my perspective) in smth else, e.g. why, when I try to
>> do
>> > the following, I see the blank response in tcpdump:
>> > [root at host1 ~]# ip a
>> > 1: lo: <LOOPBACK,UP,LOWER_UP> mtu 16436 qdisc noqueue state UNKNOWN
>> >     link/loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00
>> >     inet 127.0.0.1/8 scope host lo
>> > 2: eth0: <NO-CARRIER,BROADCAST,MULTICAST,UP> mtu 1500 qdisc pfifo_fast
>> state
>> > DOWN qlen 1000
>> >     link/ether 00:25:90:30:34:42 brd ff:ff:ff:ff:ff:ff
>> >     inet 46.../27 brd 46.182.25.159 scope global eth0
>> > 3: eth1: <BROADCAST,MULTICAST,UP,LOWER_UP> mtu 1500 qdisc pfifo_fast
>> state
>> > UP qlen 1000
>> >     link/ether 00:25:90:30:34:43 brd ff:ff:ff:ff:ff:ff
>> >     inet 192.168.1.192/24 brd 192.168.1.255 scope global eth1
>> > [root at host1 ~]# tcpdump -i eth1 'port 27004'
>> > tcpdump: verbose output suppressed, use -v or -vv for full protocol
>> decode
>> > listening on eth1, link-type EN10MB (Ethernet), capture size 65535 bytes
>> > ^Z
>> > [2]+  Stopped                 tcpdump -i eth1 'port 27004'
>> > [root at host1 ~]# bg
>> > [2]+ tcpdump -i eth1 'port 27004' &
>> >
>> > [root at host1 ~]# gluster peer probe host2
>> > Probe on localhost not needed
>> > [root at host1 ~]# fg
>> > tcpdump -i eth1 'port 27004'
>> > ^C
>> > 0 packets captured
>> > 117 packets received by filter
>> > 52 packets dropped by kernel
>> >
>> > So (by whatever reason) Gluster doesn't send a packet to anywhere
>> through
>> > ethernet.
>> >
>> > Guys, any ideas?
>> >
>> > Thanks,
>> > BR,
>> > vy
>> >
>> >
>> > On Mon, Jul 2, 2012 at 6:06 PM, Brian Candler <B.Candler at pobox.com>
>> wrote:
>> >>
>> >> On Mon, Jul 02, 2012 at 05:54:46PM +0400, Vladimir Yakovlev wrote:
>> >> >    I tried different configurations, the latest follows:
>> >> >    [root at host1 ~]# cat /etc/hosts
>> >> >    127.0.0.1   localhost localhost.localdomain localhost4
>> >> >    localhost4.localdomain4
>> >> >    127.0.0.1   host1 host1.fqdn
>> >> >    192.168.1.193   host2 host2.fqdn
>> >> >    [root at host2 ~]# cat /etc/hosts
>> >> >    127.0.0.1   localhost localhost.localdomain localhost4
>> >> >    localhost4.localdomain4
>> >> >    127.0.0.1   host2 host2.fqdn
>> >> >    192.168.1.192   host1 host1.fqdn
>> >>
>> >> Be consistent. Both machines should have:
>> >>
>> >> 127.0.0.1      localhost localhost.localdomain
>> >> 192.168.1.192  host1.fqdn host1
>> >> 192.168.1.193  host2.fqdn host2
>> >>
>> >> On host1 you should have "host1.fqdn" in /etc/hostname, and the command
>> >> "hostname" should show it. (Ditto for host2, but with "host2.fqdn" of
>> >> course)
>> >
>> >
>> >
>> > _______________________________________________
>> > Gluster-users mailing list
>> > Gluster-users at gluster.org
>> > http://gluster.org/cgi-bin/mailman/listinfo/gluster-users
>> >
>>
>
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://supercolony.gluster.org/pipermail/gluster-users/attachments/20120702/468354fa/attachment.html>