glusterfs.git/libglusterfs/src, branch v3.4.0alpha2

glusterd: Increasing throughput of synctask based mgmt ops.

2013-03-07T06:06:30+00:00

Change-Id: Ibd963f78707b157fc4c9729aa87206cfd5ecfe81
BUG: 913662
Signed-off-by: Krishnan Parthasarathi 
Reviewed-on: http://review.gluster.org/4638
Tested-by: Gluster Build System 
Reviewed-by: Jeff Darcy

synctask: support for (assymetric) counted barriers

2013-03-07T05:32:34+00:00

[Backport of Avati's patch on master - http://review.gluster.org/4558]
This patch introduces a new set of primitives:

  - synctask_barrier_init (stub)
  - synctask_barrier_waitfor (stub, count)
  - synctask_barrier_wake (stub)

Unlike pthread_barrier_t, this barrier has an explicit notion of
"waiter" and "waker". The "waiter" waits for @count number of
"wakers" to call synctask_barrier_wake() before returning. The
wait performed by the waiter via synctask_barrier_waitfor() is
co-operative in nature and yields the thread for scheduling other
synctasks in the mean time.

Intended use case:

  Eliminate excessive serialization in glusterd and allow for
concurrent RPC transactions.

  Code which are currently in this format:

---old---

  list_for_each_entry (peerinfo, peers, op_peers_list) {
          ...
          GD_SYNCOP (peerinfo->rpc, stub, rpc_cbk, ...);
  }

  ...

  int rpc_cbk (rpc, stub, ...)
  {
          ...
          __wake (stub);
  }

---old---

  Can be restructred into the format:

---new---

  synctask_barrier_init (stub);
  {
          list_for_each_entry (peerinfo, peers, op_peers_list) {
                  ...
                  rpc_submit (peerinfo->rpc, stub, rpc_cbk, ...);
                  count++;
           }
   }
   synctask_barrier_wait (stub, count);

   ...

   int rpc_cbk (rpc, stub, ...)
   {
           ...
           synctask_barrier_wake (stub);
   }

---new---

  In the above structure, from the synctask's point of view, the region
between synctask_barrier_init() and synctask_barrier_wait() are spawning
off asynchronous "threads" (or RPC) and keep count of how many such
threads have been spawned. Each of those threads are expected to make
one call to synctask_barrier_wake(). The call to synctask_barrier_wait()
makes the synctask thread co-operatively wait/sleep till @count such threads
call their wake function.

  This way, the synctask thread retains the "synchronous" flow in the code,
yet at the same time allows for asynchronous "threads" to acheive parallelism
over RPC.

Change-Id: Ie037f99b2d306b71e63e3a56353daec06fb0bf41
BUG: 913662
Signed-off-by: Anand Avati 
Signed-off-by: Krishnan Parthasarathi 
Reviewed-on: http://review.gluster.org/4636
Reviewed-by: Jeff Darcy 
Tested-by: Gluster Build System

synctask: implement setuid-like SYNCTASK_SETID()

2013-03-04T10:42:19+00:00

synctasks can now call SYNCTASK_SETID(uid,gid) to set the effective
uid/gid of the frame with which the FOP will be performed.

Once called, the uid/gid is set either till the end of the synctask
or till the next call of SYNCTASK_SETID()

Back-porting Avati's patch http://review.gluster.org/#change,4269

BUG: 884597
Change-Id: Id0569da4bb8959636881457217fe004bf30c5b9d
Signed-off-by: shishir gowda 
Reviewed-on: http://review.gluster.org/4611
Tested-by: Gluster Build System 
Reviewed-by: Jeff Darcy

libglusterfs: Fix memory leaks in fd_lk_insert_and_merge

2013-03-03T14:19:26+00:00

Change-Id: I666664895fdd7c7199797796819e652557a7ac99
BUG: 834465
Signed-off-by: Vijay Bellur 
Reviewed-on: http://review.gluster.org/4529
Reviewed-by: Raghavendra Bhat 
Tested-by: Gluster Build System 
Reviewed-by: Jeff Darcy

Better mechanism to handle memory accounting

2013-03-03T14:18:59+00:00

Memory accounting will now be enabled if:

1) Any glusterfs process is spawned with argument --mem-accounting.
2) DEBUG is defined.

Change-Id: I3345e114127a57ce61916be0e2c4e0049a4c3432
BUG: 834465
Signed-off-by: Vijay Bellur 
Reviewed-on: http://review.gluster.org/4527
Reviewed-by: Raghavendra Bhat 
Tested-by: Gluster Build System 
Reviewed-by: Jeff Darcy

glusterd: harden 'volume start' staging to check for brick dirs' presence

2013-02-09T03:13:36+00:00

PROBLEM:

When the brick directory of a volume is absent on any of the servers,
AND an attempt is made to start the volume, commit fails ONLY on the
node where the brick dir is absent, leading to a split-brain like
situation.

FIX:

Harden 'volume start' to check for the presence of brick directories
at the time of staging, thereby preventing commit failure.

Change-Id: I67faeb9afbd3aa76f08645924462db126bf7a977
BUG: 889996
Signed-off-by: Krutika Dhananjay 
Reviewed-on: http://review.gluster.org/4365
Tested-by: Gluster Build System 
Reviewed-by: Anand Avati

rpc: get hostnames of client to allow FQDN based authentication

2013-02-07T00:15:54+00:00

If FQDNs are used to authenticate clients, then from
this commit forth, the client ip(v4,6) is reverse looked up
using getnameinfo to get a hostname associated with it,
if any, thereby making FQDN-based rpc authentication possible.

Change-Id: I4c5241e7079a2560de79ca15f611e65c0b858f9b
BUG: 903553
Signed-off-by: Rajesh Amaravathi 
Reviewed-on: http://review.gluster.org/4439
Tested-by: Gluster Build System 
Reviewed-by: Anand Avati

libglusterfs/core: dump frame and callstack creation times in statedump

2013-02-06T20:18:49+00:00

Change-Id: I3772602ac264cbca490d77a0343038297faee7df
BUG: 844688
Signed-off-by: Raghavendra Bhat 
Reviewed-on: http://review.gluster.org/4087
Tested-by: Gluster Build System 
Reviewed-by: Anand Avati

glusterd,glusterfsd,libgfapi: Client op-version

2013-02-06T20:17:06+00:00

This patch introduces op-version support for glusterfs clients.
Now, a client sends its supported op-versions during the volfile fetch request
and glusterd will return the volfile only if the client can support the current
op-version of the cluster.

Change-Id: Iab1f1f1706802962bcf27058657c44e8a344d2f6
BUG: 907311
Signed-off-by: Kaushal M 
Reviewed-on: http://review.gluster.org/4247
Tested-by: Gluster Build System 
Reviewed-by: Amar Tumballi 
Reviewed-by: Anand Avati

libglusterfs: Maintain open-fd-count in inode.

2013-02-06T17:48:23+00:00

Change-Id: I643d02959f92e40f68a53baf165753ed20f8b3e0
BUG: 908146
Signed-off-by: Pranith Kumar K 
Reviewed-on: http://review.gluster.org/4468
Tested-by: Gluster Build System 
Reviewed-by: Anand Avati