[Gluster-users] [Gluster-devel] OS X porting merged

[Justin Clift]

On 03/05/2014, at 8:04 AM, Harshavardhana wrote:
> Justin,
> 
> Here you go - http://review.gluster.org/#/c/7651/


Thanks, that fixes the umount2 problem. :)

Things still aren't working correctly though.  Sig 11:

  http://fpaste.org/98915/39915131/

That's using this git commit as latest (prior to adding
your patch on top):

  commit b3031351b0e69195413f5f5b9cad2752e3eb713d
  Author: Sachin Pandit <spandit at redhat.com>
  Date:   Sat Apr 26 14:06:47 2014 +0530

    glusterd/snapshot : Copy the quota config and cksum file before taking a snapshot

Note, this is my configure argument:

  $ export MYBASE=~/local/gluster
  $ ./configure --prefix=$MYBASE --disable-systemtap --with-mountutildir=$MYBASE/sbin \
    --with-initdir=$MYBASE/etc --with-launchddir=$MYBASE/Library/LaunchDaemons \
    --localstatedir=$MYBASE/statedir

With this configure summary:

  GlusterFS configure summary
  ===========================
  FUSE client          : yes
  Infiniband verbs     : no
  epoll IO multiplex   : no
  argp-standalone      : yes
  fusermount           : no
  readline             : no (present but missing undo)
  georeplication       : no
  Linux-AIO            : no
  Enable Debug         : no
  systemtap            : no
  Block Device xlator  : no
  glupy                : no
  Use syslog           : yes
  XML output           : yes
  QEMU Block formats   : no
  Encryption xlator    : no

Not sure why it's trying to put stuff in /var/lib/gluster, but it's not
serious problem.  Will figure that out later. :)

On a side point, updating to very latest git head before applying your
patch shows a different bug first:

  $ sudo glusterd --debug
  [2014-05-03 20:49:34.684205] I [MSGID: 100030] [glusterfsd.c:2016:main] 0-glusterd: Started running glusterd version 3.5qa2 (args: glusterd --debug)
  [2014-05-03 20:49:34.684386] D [MSGID: 0] [glusterfsd.c:614:get_volfp] 0-glusterfsd: loading volume file /Users/jc/local/gluster/etc/glusterfs/glusterd.vol
  [2014-05-03 20:49:34.685497] I [glusterd.c:1228:init] 0-management: Using /var/lib/glusterd as working directory
  [2014-05-03 20:49:34.685536] E [common-utils.c:105:mkdir_p] 0-: /var is a symlink
  [2014-05-03 20:49:34.685549] C [glusterd.c:1154:glusterd_init_snap_folder] 0-management: Unable to create directory /var/run/gluster/snaps ,errno = 17
  [2014-05-03 20:49:34.685558] C [glusterd.c:1234:init] 0-management: Unable to create snap backend folder
  $

The "E [common-utils.c:105:mkdir_p] 0-: /var is a symlink" seems a bit weird.
(Sure it's a symlink, who cares? why is that an error?)  The "Unable to create
snap backend folder" seems like a bug in code merged since yesterday. :(  I
don't have time to isolate the exact commit which introduces this just yet.

Anyway, it can be worked around by  "sudo mkdir -p /var/run/gluster/snaps"
before running glusterd.  Then the sig 11 error from above happens.

+ Justin

--
Open Source and Standards @ Red Hat

twitter.com/realjustinclift