gluster-devel
[Top][All Lists]
Advanced

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: [Gluster-devel] error gf_block_unserialize_transport server: EOFfrom


From: Rob Derr
Subject: Re: [Gluster-devel] error gf_block_unserialize_transport server: EOFfrom peer
Date: Tue, 29 Jan 2008 10:59:29 -0500

Gowda,

I'm running gluster in a production environment and getting similar errors
as Rohan's.

We have 8 computers running as servers and 11 clients connecting to the
servers (including the 8 clients that are also servers).  The servers are
grouped into 4 mirrors and then unified as one large filesystem.

This is our current client spec file:

volume client1
  type protocol/client
  option transport-type tcp/client
  option remote-host 192.168.2.46
  option remote-subvolume brick
  option transport-timeout 10
end-volume

volume client2
  type protocol/client
  option transport-type tcp/client
  option remote-host 192.168.2.47
  option remote-subvolume brick
  option transport-timeout 10
end-volume

volume client3
  type protocol/client
  option transport-type tcp/client
  option remote-host 192.168.2.48
  option remote-subvolume brick
  option transport-timeout 10
end-volume

volume client4
  type protocol/client
  option transport-type tcp/client
  option remote-host 192.168.2.49
  option remote-subvolume brick
  option transport-timeout 10
end-volume

volume client5
  type protocol/client
  option transport-type tcp/client
  option remote-host 192.168.2.50
  option remote-subvolume brick
  option transport-timeout 10
end-volume

volume client6
  type protocol/client
  option transport-type tcp/client
  option remote-host 192.168.2.65
  option remote-subvolume brick
  option transport-timeout 10
end-volume

volume client7
  type protocol/client
  option transport-type tcp/client
  option remote-host 192.168.2.43
  option remote-subvolume brick
  option transport-timeout 10
end-volume

volume client8
  type protocol/client
  option transport-type tcp/client
  option remote-host 192.168.2.44
  option remote-subvolume brick
  option transport-timeout 10
end-volume


### Client Namespaves
volume client1ns
  type protocol/client
  option transport-type tcp/client
  option remote-host 192.168.2.46
  option remote-subvolume brickns
  option transport-timeout 10
end-volume

volume client2ns
  type protocol/client
  option transport-type tcp/client
  option remote-host 192.168.2.47
  option remote-subvolume brickns
  option transport-timeout 10
end-volume

volume brick-ns
  type cluster/afr
  subvolumes client1ns client2ns
end-volume

volume afr1
  type cluster/afr
  subvolumes client1 client2
end-volume

volume afr2
  type cluster/afr
  subvolumes client3 client4
end-volume

volume afr3
  type cluster/afr
  subvolumes client5 client6
end-volume

volume afr4
  type cluster/afr
  subvolumes client7 client8
end-volume

# Unify
volume bricks
  type cluster/unify
  subvolumes afr1 afr2 afr3 afr4
  option namespace brick-ns
end-volume

This is the server spec file:
volume brick
  type storage/posix                   # POSIX FS translator
  option directory /var/gluster/export        # Export this directory
end-volume

volume brickns
  type storage/posix                   # POSIX FS translator
  option directory /var/gluster/export-ns        # Export this directory
end-volume

### Add network serving capability to above brick.
volume server
  type protocol/server
  option transport-type tcp/server     # For TCP/IP transport
  subvolumes brick brickns
  option auth.ip.brick.allow 192.168.2.* # Allow access to "brick" volume
  option auth.ip.brickns.allow 192.168.2.* # Allow access to "brick" volume
end-volume

Here are the server logs from the last time the servers failed:

2008-01-29 05:22:44 E [protocol.c:254:gf_block_unserialize_transport]
server: EOF from peer (192.168.2.49:1023)
2008-01-29 05:22:44 E [posix.c:1209:posix_close] brick: pfd->dir is 0x512ff0
(not NULL) for file fd=0x50ed50
2008-01-29 05:22:44 E [posix.c:1209:posix_close] brick: pfd->dir is
0x2aaaab002610 (not NULL) for file fd=0x2aaaab000d10
2008-01-29 05:22:44 C [tcp.c:81:tcp_disconnect] server: connection
disconnected
2008-01-29 05:25:19 E [protocol.c:254:gf_block_unserialize_transport]
server: EOF from peer (192.168.2.43:1001)
2008-01-29 05:25:19 C [tcp.c:81:tcp_disconnect] server: connection
disconnected


Here are the client logs:

2008-01-29 05:22:16 C [client-protocol.c:217:call_bail] client4: bailing
transport
2008-01-29 05:22:16 E [client-protocol.c:4555:client_protocol_cleanup]
client4: forced unwinding frame type(0) op(34) address@hidden
2008-01-29 05:22:16 C [tcp.c:81:tcp_disconnect] client4: connection
disconnected
2008-01-29 05:22:36 C [client-protocol.c:217:call_bail] client4: bailing
transport
2008-01-29 05:22:36 E [client-protocol.c:4555:client_protocol_cleanup]
client4: forced unwinding frame type(0) op(34) address@hidden
2008-01-29 05:22:36 C [tcp.c:81:tcp_disconnect] client4: connection
disconnected
2008-01-29 05:22:56 C [client-protocol.c:217:call_bail] client4: bailing
transport
2008-01-29 05:22:56 E [client-protocol.c:4555:client_protocol_cleanup]
client4: forced unwinding frame type(0) op(34) address@hidden
2008-01-29 05:22:56 C [tcp.c:81:tcp_disconnect] client4: connection
disconnected
2008-01-29 05:23:16 C [client-protocol.c:217:call_bail] client4: bailing
transport
2008-01-29 05:23:16 E [client-protocol.c:4555:client_protocol_cleanup]
client4: forced unwinding frame type(0) op(34) address@hidden
2008-01-29 05:23:16 C [tcp.c:81:tcp_disconnect] client4: connection
disconnected
2008-01-29 05:24:14 C [client-protocol.c:217:call_bail] client4: bailing
transport
2008-01-29 05:24:14 E [client-protocol.c:4555:client_protocol_cleanup]
client4: forced unwinding frame type(0) op(34) address@hidden
2008-01-29 05:24:14 C [tcp.c:81:tcp_disconnect] client4: connection
disconnected
2008-01-29 05:25:19 C [client-protocol.c:217:call_bail] client4: bailing
transport
2008-01-29 05:25:19 E [client-protocol.c:4555:client_protocol_cleanup]
client4: forced unwinding frame type(0) op(34) address@hidden
2008-01-29 05:25:19 C [tcp.c:81:tcp_disconnect] client4: connection
disconnected
2008-01-29 05:25:37 C [client-protocol.c:217:call_bail] client4: bailing
transport
2008-01-29 05:25:37 E [client-protocol.c:4555:client_protocol_cleanup]
client4: forced unwinding frame type(0) op(34) address@hidden
2008-01-29 05:25:37 E [client-protocol.c:4555:client_protocol_cleanup]
client4: forced unwinding frame type(0) op(34) address@hidden
2008-01-29 05:25:37 C [tcp.c:81:tcp_disconnect] client4: connection
disconnected
2008-01-29 05:26:10 C [client-protocol.c:217:call_bail] client4: bailing
transport
2008-01-29 05:26:10 E [client-protocol.c:4555:client_protocol_cleanup]
client4: forced unwinding frame type(0) op(34) address@hidden
2008-01-29 05:26:10 C [tcp.c:81:tcp_disconnect] client4: connection
disconnected
2008-01-29 05:27:20 C [client-protocol.c:217:call_bail] client4: bailing
transport
2008-01-29 05:27:20 E [client-protocol.c:4555:client_protocol_cleanup]
client4: forced unwinding frame type(0) op(34) address@hidden
2008-01-29 05:27:20 C [tcp.c:81:tcp_disconnect] client4: connection
disconnected
2008-01-29 05:27:59 E [tcp-client.c:171:tcp_connect] client4: non-blocking
connect() returned: 113 (No route to host)
2008-01-29 10:16:24 E [fuse-bridge.c:431:fuse_entry_cbk] glusterfs-fuse:
6336: /weatherflow_wp/cache/wp-cache-8015965cebd529c3433e331c330e193b.meta=>
-1 (2)
2008-01-29 10:16:24 E [fuse-bridge.c:431:fuse_entry_cbk] glusterfs-fuse:
6336: /weatherflow_wp/cache/wp-cache-8015965cebd529c3433e331c330e193b.meta=>
-1 (2)
2008-01-29 10:16:41 E [fuse-bridge.c:431:fuse_entry_cbk] glusterfs-fuse:
6442: /weatherflow_wp/cache/wp-cache-bb98ceeeb4ca784b406a1aedd2df7bc0.meta=>
-1 (2)


We are running fuse 2.6.5-2.fc5 and glusterfs--mainline--2.5 patch-643

The problem starts around 5:00am when our cleanup scripts run.  It was
failing everyday when updatedb ran across the cluster on every server, but
since I setup updatedb to ignore gluster mounts the system is able to stay
up much longer now.


On Jan 29, 2008 8:19 AM, Basavanagowda Kanur <address@hidden> wrote:

> Rohan,
>  Can you please send us the server logs and also the volume spec files of
> both server and client? That will help us find out the exact reason for
> this
> problem.
>
> --
> Gowda
>
>


reply via email to

[Prev in Thread] Current Thread [Next in Thread]