[Gluster-devel] AFR over NUFA with 1.3.5

[Thierry EXCOFFIER]

Hello.

Does AFR can work over UNIFY ?

I tried with NUFA and RANDOM scheduler but there is
a core dump when starting the 'glusterfs' mount.

I attached the log with backtrace and the configuration file.

A simple AFR on two bricks works perfectly for me.


Thanks to all the glusterfs developpers for this great work.
Thierry Excoffier.
-- 
Thierry.EXCOFFIER@@liris.univ-lyon1.fr//   GO: Oriental Strategy Game
LIRIS/UFR Informatique, bat710       //DYLAN: Programming Language
Universite Claude Bernard LYON 1  __// UNIX: Standard Operating System
43 Bd du 11 novembre 1918       //
69622 VILLEURBANNE, FRANCE     //WWW: http://www710.univ-lyon1.fr/~exco

2007-10-14 22:35:55 D [glusterfs.c:131:get_spec_fp] glusterfs: loading spec 
from /etc/glusterfs/glusterfs-client.vol
2007-10-14 22:35:55 W [fuse-bridge.c:2104:fuse_transport_notify] 
glusterfs-fuse: Ignoring notify event 4
2007-10-14 22:35:55 D [spec.y:116:new_section] libglusterfs/parser: New node 
for 'namespace'
2007-10-14 22:35:55 D [spec.y:132:section_type] libglusterfs/parser: 
Type:namespace:storage/posix
2007-10-14 22:35:55 D [xlator.c:102:xlator_set_type] libglusterfs/xlator: 
attempt to load type storage/posix
2007-10-14 22:35:55 D [xlator.c:109:xlator_set_type] libglusterfs/xlator: 
attempt to load file //lib/glusterfs/1.3.5/xlator/storage/posix.so
2007-10-14 22:35:55 D [spec.y:152:section_option] libglusterfs/parser: 
Option:namespace:directory:/tmp/namespace
2007-10-14 22:35:55 D [spec.y:216:section_end] libglusterfs/parser: 
end:namespace
2007-10-14 22:35:55 D [spec.y:116:new_section] libglusterfs/parser: New node 
for 'big_brick'
2007-10-14 22:35:55 D [spec.y:132:section_type] libglusterfs/parser: 
Type:big_brick:protocol/client
2007-10-14 22:35:55 D [xlator.c:102:xlator_set_type] libglusterfs/xlator: 
attempt to load type protocol/client
2007-10-14 22:35:55 D [xlator.c:109:xlator_set_type] libglusterfs/xlator: 
attempt to load file //lib/glusterfs/1.3.5/xlator/protocol/client.so
2007-10-14 22:35:55 D [spec.y:152:section_option] libglusterfs/parser: 
Option:big_brick:transport-type:tcp/client
2007-10-14 22:35:55 D [spec.y:152:section_option] libglusterfs/parser: 
Option:big_brick:remote-host:192.168.0.8
2007-10-14 22:35:55 D [spec.y:152:section_option] libglusterfs/parser: 
Option:big_brick:remote-subvolume:brick
2007-10-14 22:35:55 D [spec.y:152:section_option] libglusterfs/parser: 
Option:big_brick:transport-timeout:5
2007-10-14 22:35:55 D [spec.y:216:section_end] libglusterfs/parser: 
end:big_brick
2007-10-14 22:35:55 D [spec.y:116:new_section] libglusterfs/parser: New node 
for 'small_brick_1'
2007-10-14 22:35:55 D [spec.y:132:section_type] libglusterfs/parser: 
Type:small_brick_1:protocol/client
2007-10-14 22:35:55 D [xlator.c:102:xlator_set_type] libglusterfs/xlator: 
attempt to load type protocol/client
2007-10-14 22:35:55 D [xlator.c:109:xlator_set_type] libglusterfs/xlator: 
attempt to load file //lib/glusterfs/1.3.5/xlator/protocol/client.so
2007-10-14 22:35:55 D [spec.y:152:section_option] libglusterfs/parser: 
Option:small_brick_1:transport-type:tcp/client
2007-10-14 22:35:55 D [spec.y:152:section_option] libglusterfs/parser: 
Option:small_brick_1:remote-host:192.168.0.7
2007-10-14 22:35:55 D [spec.y:152:section_option] libglusterfs/parser: 
Option:small_brick_1:remote-subvolume:brick
2007-10-14 22:35:55 D [spec.y:152:section_option] libglusterfs/parser: 
Option:small_brick_1:transport-timeout:5
2007-10-14 22:35:55 D [spec.y:216:section_end] libglusterfs/parser: 
end:small_brick_1
2007-10-14 22:35:55 D [spec.y:116:new_section] libglusterfs/parser: New node 
for 'small_brick_2'
2007-10-14 22:35:55 D [spec.y:132:section_type] libglusterfs/parser: 
Type:small_brick_2:protocol/client
2007-10-14 22:35:55 D [xlator.c:102:xlator_set_type] libglusterfs/xlator: 
attempt to load type protocol/client
2007-10-14 22:35:55 D [xlator.c:109:xlator_set_type] libglusterfs/xlator: 
attempt to load file //lib/glusterfs/1.3.5/xlator/protocol/client.so
2007-10-14 22:35:55 D [spec.y:152:section_option] libglusterfs/parser: 
Option:small_brick_2:transport-type:tcp/client
2007-10-14 22:35:55 D [spec.y:152:section_option] libglusterfs/parser: 
Option:small_brick_2:remote-host:192.168.0.1
2007-10-14 22:35:55 D [spec.y:152:section_option] libglusterfs/parser: 
Option:small_brick_2:remote-subvolume:brick
2007-10-14 22:35:55 D [spec.y:152:section_option] libglusterfs/parser: 
Option:small_brick_2:transport-timeout:5
2007-10-14 22:35:55 D [spec.y:216:section_end] libglusterfs/parser: 
end:small_brick_2
2007-10-14 22:35:55 D [spec.y:116:new_section] libglusterfs/parser: New node 
for 'nufa'
2007-10-14 22:35:55 D [spec.y:132:section_type] libglusterfs/parser: 
Type:nufa:cluster/unify
2007-10-14 22:35:55 D [xlator.c:102:xlator_set_type] libglusterfs/xlator: 
attempt to load type cluster/unify
2007-10-14 22:35:55 D [xlator.c:109:xlator_set_type] libglusterfs/xlator: 
attempt to load file //lib/glusterfs/1.3.5/xlator/cluster/unify.so
2007-10-14 22:35:55 D [spec.y:201:section_sub] liglusterfs/parser: 
child:nufa->small_brick_1
2007-10-14 22:35:55 D [spec.y:201:section_sub] liglusterfs/parser: 
child:nufa->small_brick_2
2007-10-14 22:35:55 D [spec.y:152:section_option] libglusterfs/parser: 
Option:nufa:scheduler:nufa
2007-10-14 22:35:55 D [spec.y:152:section_option] libglusterfs/parser: 
Option:nufa:nufa.local-volume-name:small_brick_2
2007-10-14 22:35:55 D [spec.y:152:section_option] libglusterfs/parser: 
Option:nufa:nufa.limits.min-free-disk:5%
2007-10-14 22:35:55 D [spec.y:152:section_option] libglusterfs/parser: 
Option:nufa:namespace:namespace
2007-10-14 22:35:55 D [spec.y:216:section_end] libglusterfs/parser: end:nufa
2007-10-14 22:35:55 W [inode.c:1099:inode_table_new] fuse: creating new inode 
table with lru_limit=1024, sizeof(inode_t)=96
2007-10-14 22:35:55 D [inode.c:559:__create_inode] fuse/inode: create inode(1)
2007-10-14 22:35:55 D [inode.c:351:__active_inode] fuse/inode: activating 
inode(1), lru=0/1024
2007-10-14 22:35:55 D [client-protocol.c:4549:init] small_brick_1: missing 
'inode-lru-limit'. defaulting to 1000
2007-10-14 22:35:55 D [client-protocol.c:4562:init] small_brick_1: setting 
transport-timeout to 5
2007-10-14 22:35:55 D [transport.c:83:transport_load] libglusterfs/transport: 
attempt to load type tcp/client
2007-10-14 22:35:55 D [transport.c:88:transport_load] libglusterfs/transport: 
attempt to load file //lib/glusterfs/1.3.5/transport/tcp/client.so
2007-10-14 22:35:55 D [client-protocol.c:4549:init] small_brick_2: missing 
'inode-lru-limit'. defaulting to 1000
2007-10-14 22:35:55 D [client-protocol.c:4562:init] small_brick_2: setting 
transport-timeout to 5
2007-10-14 22:35:55 D [transport.c:83:transport_load] libglusterfs/transport: 
attempt to load type tcp/client
2007-10-14 22:35:55 D [transport.c:88:transport_load] libglusterfs/transport: 
attempt to load file //lib/glusterfs/1.3.5/transport/tcp/client.so
2007-10-14 22:35:55 D [unify.c:3872:init] nufa: namespace node specified as 
namespace
2007-10-14 22:35:55 D [scheduler.c:36:get_scheduler] libglusterfs/scheduler: 
attempt to load file nufa.so

2007-10-14 22:35:55 D [unify.c:3890:init] nufa: Child node count is 2
2007-10-14 22:35:55 D [client-protocol.c:4847:notify] small_brick_1: got 
GF_EVENT_PARENT_UP, attempting connect on transport
2007-10-14 22:35:55 D [client-protocol.c:4847:notify] small_brick_2: got 
GF_EVENT_PARENT_UP, attempting connect on transport
2007-10-14 22:35:55 D [client-protocol.c:4294:client_protocol_reconnect] 
small_brick_1: attempting reconnect
2007-10-14 22:35:55 D [tcp-client.c:70:tcp_connect] small_brick_1: socket fd = 4
2007-10-14 22:35:55 D [tcp-client.c:88:tcp_connect] small_brick_1: finalized on 
port `1023'
2007-10-14 22:35:55 D [tcp-client.c:109:tcp_connect] small_brick_1: defaulting 
remote-port to 6996
2007-10-14 22:35:55 D [common-utils.c:161:gf_resolve_ip] resolver: DNS cache 
not present, freshly probing hostname: 192.168.0.7
2007-10-14 22:35:55 D [common-utils.c:186:gf_resolve_ip] resolver: returning 
IP:192.168.0.7[0] for hostname: 192.168.0.7
2007-10-14 22:35:55 D [common-utils.c:194:gf_resolve_ip] resolver: flushing DNS 
cache
2007-10-14 22:35:55 D [tcp-client.c:142:tcp_connect] small_brick_1: connect on 
4 in progress (non-blocking)
2007-10-14 22:35:55 D [tcp-client.c:179:tcp_connect] small_brick_1: connection 
on 4 still in progress - try later
2007-10-14 22:35:55 D [client-protocol.c:4294:client_protocol_reconnect] 
small_brick_2: attempting reconnect
2007-10-14 22:35:55 D [tcp-client.c:70:tcp_connect] small_brick_2: socket fd = 7
2007-10-14 22:35:55 D [tcp-client.c:88:tcp_connect] small_brick_2: finalized on 
port `1022'
2007-10-14 22:35:55 D [tcp-client.c:109:tcp_connect] small_brick_2: defaulting 
remote-port to 6996
2007-10-14 22:35:55 D [common-utils.c:161:gf_resolve_ip] resolver: DNS cache 
not present, freshly probing hostname: 192.168.0.1
2007-10-14 22:35:55 D [common-utils.c:186:gf_resolve_ip] resolver: returning 
IP:192.168.0.1[0] for hostname: 192.168.0.1
2007-10-14 22:35:55 D [common-utils.c:194:gf_resolve_ip] resolver: flushing DNS 
cache
2007-10-14 22:35:55 D [tcp-client.c:142:tcp_connect] small_brick_2: connect on 
7 in progress (non-blocking)
2007-10-14 22:35:55 D [tcp-client.c:186:tcp_connect] small_brick_2: connection 
on 7 success
2007-10-14 22:35:55 D [client-protocol.c:4869:notify] small_brick_2: got 
GF_EVENT_CHILD_UP
2007-10-14 22:35:55 D [client-protocol.c:4627:client_protocol_handshake_reply] 
small_brick_2: reply frame has callid: 424242
2007-10-14 22:35:55 D [client-protocol.c:4661:client_protocol_handshake_reply] 
small_brick_2: SETVOLUME on remote-host succeeded
2007-10-14 22:35:56 D [client-protocol.c:4294:client_protocol_reconnect] 
small_brick_1: attempting reconnect
2007-10-14 22:35:56 D [tcp-client.c:179:tcp_connect] small_brick_1: connection 
on 4 still in progress - try later
2007-10-14 22:35:56 D [client-protocol.c:4301:client_protocol_reconnect] 
small_brick_2: breaking reconnect chain
2007-10-14 22:35:58 D [client-protocol.c:4294:client_protocol_reconnect] 
small_brick_1: attempting reconnect
2007-10-14 22:35:58 E [tcp-client.c:171:tcp_connect] small_brick_1: 
non-blocking connect() returned: 113 (No route to host)
2007-10-14 22:36:00 D [tcp-client.c:70:tcp_connect] small_brick_1: socket fd = 4
2007-10-14 22:36:00 D [tcp-client.c:88:tcp_connect] small_brick_1: finalized on 
port `1023'
2007-10-14 22:36:00 D [tcp-client.c:109:tcp_connect] small_brick_1: defaulting 
remote-port to 6996
2007-10-14 22:36:00 D [common-utils.c:161:gf_resolve_ip] resolver: DNS cache 
not present, freshly probing hostname: 192.168.0.7
2007-10-14 22:36:00 D [common-utils.c:186:gf_resolve_ip] resolver: returning 
IP:192.168.0.7[0] for hostname: 192.168.0.7
2007-10-14 22:36:00 D [common-utils.c:194:gf_resolve_ip] resolver: flushing DNS 
cache
2007-10-14 22:36:00 D [tcp-client.c:142:tcp_connect] small_brick_1: connect on 
4 in progress (non-blocking)
2007-10-14 22:36:01 D [tcp-client.c:179:tcp_connect] small_brick_1: connection 
on 4 still in progress - try later
2007-10-14 22:36:01 W [client-protocol.c:344:client_protocol_xfer] 
small_brick_1: not connected at the moment to submit frame type(0) op(34)
2007-10-14 22:36:01 W [unify-self-heal.c:261:unify_sh_readdir_cbk] nufa: found 
mismatch in mode/uid/gid for .
2007-10-14 22:36:01 D [client-protocol.c:4294:client_protocol_reconnect] 
small_brick_1: attempting reconnect
2007-10-14 22:36:01 D [tcp-client.c:179:tcp_connect] small_brick_1: connection 
on 4 still in progress - try later
2007-10-14 22:36:06 D [client-protocol.c:4294:client_protocol_reconnect] 
small_brick_1: attempting reconnect
2007-10-14 22:36:06 E [tcp-client.c:171:tcp_connect] small_brick_1: 
non-blocking connect() returned: 113 (No route to host)
2007-10-14 22:36:14 D [client-protocol.c:4294:client_protocol_reconnect] 
small_brick_1: attempting reconnect
2007-10-14 22:36:14 D [tcp-client.c:70:tcp_connect] small_brick_1: socket fd = 4
2007-10-14 22:36:14 D [tcp-client.c:88:tcp_connect] small_brick_1: finalized on 
port `1023'
2007-10-14 22:36:14 D [tcp-client.c:109:tcp_connect] small_brick_1: defaulting 
remote-port to 6996
2007-10-14 22:36:14 D [common-utils.c:161:gf_resolve_ip] resolver: DNS cache 
not present, freshly probing hostname: 192.168.0.7
2007-10-14 22:36:14 D [common-utils.c:186:gf_resolve_ip] resolver: returning 
IP:192.168.0.7[0] for hostname: 192.168.0.7
2007-10-14 22:36:14 D [common-utils.c:194:gf_resolve_ip] resolver: flushing DNS 
cache
2007-10-14 22:36:14 D [tcp-client.c:142:tcp_connect] small_brick_1: connect on 
4 in progress (non-blocking)
2007-10-14 22:36:14 D [tcp-client.c:179:tcp_connect] small_brick_1: connection 
on 4 still in progress - try later
2007-10-14 22:36:27 D [client-protocol.c:4294:client_protocol_reconnect] 
small_brick_1: attempting reconnect
2007-10-14 22:36:27 E [tcp-client.c:171:tcp_connect] small_brick_1: 
non-blocking connect() returned: 113 (No route to host)
2007-10-14 22:36:48 D [client-protocol.c:4294:client_protocol_reconnect] 
small_brick_1: attempting reconnect
2007-10-14 22:36:48 D [tcp-client.c:70:tcp_connect] small_brick_1: socket fd = 4
2007-10-14 22:36:48 D [tcp-client.c:88:tcp_connect] small_brick_1: finalized on 
port `1023'
2007-10-14 22:36:48 D [tcp-client.c:109:tcp_connect] small_brick_1: defaulting 
remote-port to 6996
2007-10-14 22:36:48 D [common-utils.c:161:gf_resolve_ip] resolver: DNS cache 
not present, freshly probing hostname: 192.168.0.7
2007-10-14 22:36:48 D [common-utils.c:186:gf_resolve_ip] resolver: returning 
IP:192.168.0.7[0] for hostname: 192.168.0.7
2007-10-14 22:36:48 D [common-utils.c:194:gf_resolve_ip] resolver: flushing DNS 
cache
2007-10-14 22:36:48 D [tcp-client.c:142:tcp_connect] small_brick_1: connect on 
4 in progress (non-blocking)
2007-10-14 22:36:49 D [tcp-client.c:179:tcp_connect] small_brick_1: connection 
on 4 still in progress - try later
2007-10-14 22:37:23 D [client-protocol.c:4294:client_protocol_reconnect] 
small_brick_1: attempting reconnect
2007-10-14 22:37:23 E [tcp-client.c:171:tcp_connect] small_brick_1: 
non-blocking connect() returned: 113 (No route to host)
2007-10-14 22:38:18 D [client-protocol.c:4294:client_protocol_reconnect] 
small_brick_1: attempting reconnect
2007-10-14 22:38:18 D [tcp-client.c:70:tcp_connect] small_brick_1: socket fd = 4
2007-10-14 22:38:18 D [tcp-client.c:88:tcp_connect] small_brick_1: finalized on 
port `1023'
2007-10-14 22:38:18 D [tcp-client.c:109:tcp_connect] small_brick_1: defaulting 
remote-port to 6996
2007-10-14 22:38:18 D [common-utils.c:161:gf_resolve_ip] resolver: DNS cache 
not present, freshly probing hostname: 192.168.0.7
2007-10-14 22:38:18 D [common-utils.c:186:gf_resolve_ip] resolver: returning 
IP:192.168.0.7[0] for hostname: 192.168.0.7
2007-10-14 22:38:18 D [common-utils.c:194:gf_resolve_ip] resolver: flushing DNS 
cache
2007-10-14 22:38:18 D [tcp-client.c:142:tcp_connect] small_brick_1: connect on 
4 in progress (non-blocking)
2007-10-14 22:38:18 D [tcp-client.c:179:tcp_connect] small_brick_1: connection 
on 4 still in progress - try later
2007-10-14 22:39:47 D [client-protocol.c:4294:client_protocol_reconnect] 
small_brick_1: attempting reconnect
2007-10-14 22:39:47 E [tcp-client.c:171:tcp_connect] small_brick_1: 
non-blocking connect() returned: 113 (No route to host)
2007-10-14 22:42:11 D [client-protocol.c:4294:client_protocol_reconnect] 
small_brick_1: attempting reconnect
2007-10-14 22:42:11 D [tcp-client.c:70:tcp_connect] small_brick_1: socket fd = 4
2007-10-14 22:42:11 D [tcp-client.c:88:tcp_connect] small_brick_1: finalized on 
port `1023'
2007-10-14 22:42:11 D [tcp-client.c:109:tcp_connect] small_brick_1: defaulting 
remote-port to 6996
2007-10-14 22:42:11 D [common-utils.c:161:gf_resolve_ip] resolver: DNS cache 
not present, freshly probing hostname: 192.168.0.7
2007-10-14 22:42:11 D [common-utils.c:186:gf_resolve_ip] resolver: returning 
IP:192.168.0.7[0] for hostname: 192.168.0.7
2007-10-14 22:42:11 D [common-utils.c:194:gf_resolve_ip] resolver: flushing DNS 
cache
2007-10-14 22:42:11 D [tcp-client.c:142:tcp_connect] small_brick_1: connect on 
4 in progress (non-blocking)
2007-10-14 22:42:11 D [tcp-client.c:179:tcp_connect] small_brick_1: connection 
on 4 still in progress - try later

volume namespace
  type storage/posix
  option directory /tmp/namespace
end-volume

volume big_brick
  type protocol/client
  option transport-type tcp/client
  option remote-host 192.168.0.8
  option remote-subvolume brick
  option transport-timeout 5
end-volume

volume small_brick_1
  type protocol/client
  option transport-type tcp/client
  option remote-host 192.168.0.7
  option remote-subvolume brick
  option transport-timeout 5
end-volume

volume small_brick_2
  type protocol/client
  option transport-type tcp/client
  option remote-host 192.168.0.1
  option remote-subvolume brick
  option transport-timeout 5
end-volume

volume nufa
  type cluster/unify
  subvolumes small_brick_1 small_brick_2
  option scheduler nufa # or random
  option nufa.local-volume-name small_brick_2
  option nufa.limits.min-free-disk 5%
  option namespace namespace
end-volume