[Gluster-users] vmware file over glusterfs

[Davide Cazzola]

Hi all,
    we are running stress test and virtual machine (VMWARE) on expensive
hardware, after trying glusterfs on cheaper hardware.
Our platform is as follow:
- 2 node (SunFire X4150), 
- OS: Debian Lenny , kernel 2.6.24, glusterfs 1.3.12, fuse-2.7.3glfs10
(from your repository)
Glusterfs and fuse are compiled on both nodes; setup provides a
server/client configuration with afr and 2 bricks per node
We have 2 NIC per node directly connected for improved reliability.
the config file is reported below.

>From log files of both nodes we have a lot of next messagges:
"
...
2008-09-26 06:25:21 E [unify.c:182:unify_lookup_cbk] unify: afr2
returned 2
2008-09-26 06:25:21 E [unify.c:265:unify_lookup_cbk] unify: Revalidate
failed for /virtual_machines/vm-debug-Debian/vmware-0.log
2008-09-26 06:25:21 E [fuse-bridge.c:468:fuse_entry_cbk] glusterfs-fuse:
13393848: (34) /virtual_machines/vm-debug-Debian/vmware-0.log => -1 (2)
2008-09-26 06:25:21 W [fuse-bridge.c:411:fuse_entry_cbk] glusterfs-fuse:
13393849: (34) /virtual_machines/vm-debug-Debian/vmware.log => 101514846
Rehashing because st_nlink less than dentry maps
...
"
The alert refers to many files.
I would like to ask you what is the meaning of this log.
Any help is really appreciated.
Kind regards,
Davide

------------------------------ Config file
--------------------------------
# dataspace on storage1
volume ds1
  type storage/posix
  option directory /export/glusterfs1
end-volume

# dataspace on storage2
volume ds2
  type storage/posix
  option directory /export/glusterfs2
end-volume

# namespace for unify
volume ns1
  type storage/posix
  option directory /var/glusterfs/ns
end-volume

# posix locks
volume ds-locks1
  type features/posix-locks
  subvolumes ds1
end-volume

# posix locks
volume ds-locks2
  type features/posix-locks
  subvolumes ds2
end-volume

volume ds-threads1
  type performance/io-threads
  option thread-count 4
  option cache-size 128MB
  subvolumes ds-locks1
end-volume

volume ds-threads2
  type performance/io-threads
  option thread-count 4
  option cache-size 128MB
  subvolumes ds-locks2
end-volume

volume server
  type protocol/server
  option transport-type tcp/server
  subvolumes ds-threads1 ds-threads2 ns1
  # storage network access only
  option auth.ip.ds-threads1.allow *
  option auth.ip.ds-threads2.allow *
  option auth.ip.ns1.allow *
#  option auth.ip.gfs.allow *
end-volume


# dataspace 1 on the other node
volume ds3
  type protocol/client
  option transport-type tcp/client
  option remote-host 192.168.1.2      # storage network
  option remote-subvolume ds-threads1
#  option transport-timeout 10           # value in seconds; it should
be set relatively low
end-volume

# dataspace 2 on the other node
volume ds4
  type protocol/client
  option transport-type tcp/client
  option remote-host 192.168.1.2      # storage network
  option remote-subvolume ds-threads2
#  option transport-timeout 10           # value in seconds; it should
be set relatively low
end-volume

# namespace 2 on the other node
volume ns2
  type protocol/client
  option transport-type tcp/client
  option remote-host 192.168.1.2      # storage network
  option remote-subvolume ns1
#  option transport-timeout 10           # value in seconds; it should
be set relatively low
end-volume


# automatic file replication translator for dataspace
volume afr1
  type cluster/afr
  subvolumes ds-threads1 ds3         # local and remote dataspaces
end-volume

# automatic file replication translator for dataspace
volume afr2
  type cluster/afr
  subvolumes ds-threads2 ds4         # local and remote dataspaces
end-volume

# automatic file replication translator for dataspace
volume afr-ns
  type cluster/afr
  subvolumes ns1 ns2         # local and remote dataspaces
end-volume

volume unify
  type cluster/unify
  subvolumes afr1 afr2
  option namespace afr-ns # should be a node which is not present in
'subvolumes'
  option scheduler rr    # simple round-robin scheduler
end-volume

volume writebehind
  type performance/write-behind
  option aggregate-size 128kB
  subvolumes unify
end-volume

volume readahead
  type performance/read-ahead
  option page-size 64kB
  option page-count 16
  subvolumes writebehind
end-volume

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://supercolony.gluster.org/pipermail/gluster-users/attachments/20080926/0d3a9612/attachment.html>