glusterfs.git/xlators/mgmt/glusterd/src/glusterd-handshake.c, branch v3.5.5

rpcsvc: Validate RPC procedure number before fetch

2014-07-08T10:36:18+00:00

While accessing the procedures of given RPC program in,
rpcsvc_get_program_vector_sizer(), It was not checking boundary
conditions which would cause buffer overflow and subsequently SEGV.

Make sure rpcsvc_actor_t arrays have numactors number of actors.

FIX:
Validate the RPC procedure number before fetching the actor.

Upstream main review: http://review.gluster.org/7726

BUG: 1096020

Change-Id: Iaf207ee976cb56fa9a554ec82c9eab36d3b289ed
Signed-off-by: Santosh Kumar Pradhan 
Reviewed-on: http://review.gluster.org/8228
Tested-by: Gluster Build System 
Reviewed-by: Niels de Vos

libgfapi: Added support to fetch volume info from glusterd and store in glfs object.

2014-05-12T15:14:52+00:00

Defined new APIs in the libgfapi module, given a glfs object,
 * to send handshake RPC call to glusterd process to fetch UUID of the volume
 * store it in the glusterfs_context linked to the glfs object.
 * to parse UUID from its cannonical string format into 16-byte array
   before sending it to the libgfapi users.

Defined a RPC call in glusterd which can be used to query volume related
info by other processes using 'clnt_handshake_procs'.

Note - Currently this RPC call to glusterd process is used only to fetch UUID.
But it can be extended to get other volume related structures as well.

In addition to the above, defined a new variable to keep track of such handshake
RPCs still in progress to make sure all the corresponding RPC callbacks have been
processed before libgfapi returns the glfs object initialized.

Also bumping up the GFAPI current version number since there is a new API
"glfs_get_volume_id" defined and exposed by libgfapi as part of these changes.

Cherry picked from commit 5adb10b9ac1c634334f29732e062b12d747ae8c5:
> Change-Id: I303f76d7177d32d25bdb301b1dbcf5cd73f42807
> BUG: 1095775
> Signed-off-by: Soumya Koduri 
> Reviewed-on: http://review.gluster.org/7218
> Reviewed-by: Anand Avati 
> Reviewed-by: Harshavardhana 
> Tested-by: Gluster Build System 
> Reviewed-by: Vijay Bellur 

This change differs a little from the patch in the master branch:
- libgfapi in 3.5 does not have glfs_get_volfile(), so there were some
  merge conflicts resolved,
- libgfapi in 3.5 is not versioned, the configure.ac changes related to
  the versioning have been skipped,
- in the master branch only the XDR .x files are available, release-3.5
  requires a manual re-generation of the relates .h and .c files.

Change-Id: I52c32d0e69a52a7f4285f74164bca6fd83c4f3b3
BUG: 1095775
Signed-off-by: Soumya Koduri 
Reviewed-on: http://review.gluster.org/7741
Reviewed-by: Niels de Vos 
Tested-by: Niels de Vos

cli,glusterd: Changes to cli-glusterd communication

2013-10-17T18:26:45+00:00

Glusterd changes:
With this patch, glusterd creates a socket file in
DATADIR/run/glusterd.socket , and listen on it for cli requests. It
listens for 2 rpc programs on the socket file,
- The glusterd cli rpc program, for all cli commands
- A reduced glusterd handshake program, just for the 'system:: getspec'
  command

The location of the socket file can be changed with the glusterd option
'glusterd-sockfile'.

To retain compatibility with the '--remote-host' cli option, glusterd
also listens for the cli requests on port 24007. But, for the sake of
security, it listens using a reduced cli rpc program on the port. The
reduced rpc program only contains read-only procs used for 'volume
(info|list|status)', 'peer status' and 'system:: getwd' cli commands.

CLI changes:
The gluster cli now uses the glusterd socket file for communicating with
glusterd by default. A new option '--gluster-sock' has been added to
allow specifying the sockfile used to connect. Using the '--remote-host'
option will make cli connect to the given host & port.

Tests changes:
cluster.rc has been modified to make use of socket files and use
different log files for each glusterd.
Some of the tests using cluster.rc have been fixed.

Change-Id: Iaf24bc22f42f8014a5fa300ce37c7fc9b1b92b53
BUG: 980754
Signed-off-by: Kaushal M 
Reviewed-on: http://review.gluster.org/5280
Tested-by: Gluster Build System 
Reviewed-by: Krishnan Parthasarathi 
Reviewed-by: Vijay Bellur

glusterd: Allow bumping down a peer's op-version during probe

2013-09-10T16:07:56+00:00

Earlier, a peer running a higher op-version couldn't be probed into a
cluster running at a lower op-version. This created issues when trying
to expand an upgraded cluster. This patch changes this behaviour.

The cluster no longer rejects a peer being probed if its op-version is
higher than the cluster op-version. The peer will reduce its op-version
if it doesn't have any volumes. If the peer contains volumes and needs
to reduce its op-version, it fails the handshake and the probe fails.

Change-Id: I12c6c873922799e1557b7184e956baea643d0dea
BUG: 1005038
Signed-off-by: Kaushal M 
Reviewed-on: http://review.gluster.org/5715
Tested-by: Gluster Build System 
Reviewed-by: Vijay Bellur

glusterd/common-utils: move hostname helper functions to common-utils

2013-07-04T14:29:22+00:00

Change-Id: If47e209cb61ea0eb74ee2d6ef9e9342b2d6ee13a
BUG: 980838
Signed-off-by: Krishnan Parthasarathi 
Reviewed-on: http://review.gluster.org/5261
Tested-by: Gluster Build System 
Reviewed-by: Vijay Bellur

rpc: duplicate request cache for nfs

2013-06-21T17:33:49+00:00

Duplicate request cache provides a mechanism for detecting
duplicate rpc requests from clients. DRC caches replies
and on duplicate requests, sends the cached reply instead of
re-processing the request.

Change-Id: I3d62a6c4aa86c92bf61f1038ca62a1a46bf1c303
BUG: 847624
Signed-off-by: Rajesh Amaravathi 
Reviewed-on: http://review.gluster.org/4049
Reviewed-by: Rajesh Joseph 
Tested-by: Gluster Build System 
Reviewed-by: Anand Avati

glusterd: Log peer op status at the appropriate time

2013-06-19T05:15:12+00:00

Change-Id: Ia8e1af082078f2f791708ba4faa4992bf291dd6e
BUG: 961339
Signed-off-by: Krutika Dhananjay 
Reviewed-on: http://review.gluster.org/5023
Reviewed-by: Amar Tumballi 
Tested-by: Gluster Build System 
Reviewed-by: Vijay Bellur

glusterd: Introduce volume op-versions

2013-04-26T08:45:06+00:00

Each volume is now associated with two op-versions,
 * op_version - the op-version of the highest op-versioned feature enabled
 * client_op_version - the op-version of the highest op-versioned feature
   enabled which affects the clients only.

These two op-versions are generated dynamically and kept updated during
runtime. Glusterd now uses the respective volumes' client-op-version during
getspec requests.

To achieve the above a new field in the vme table is introduced,
client_option, this boolean field tells if the option is a client side
option.

Change-Id: I12c83b1dd29ab506026efd50d448cebbcee53c27
BUG: 907311
Signed-off-by: Kaushal M 
Reviewed-on: http://review.gluster.org/4584
Reviewed-by: Krishnan Parthasarathi 
Tested-by: Gluster Build System 
Reviewed-by: Vijay Bellur

glusterd: big lock - a coarse-grained locking to prevent races

2013-04-12T20:47:46+00:00

There are primarily three lists that are part of glusterd process,
that are concurrently accessed. Namely, priv->volumes, priv->peers
and volinfo->bricks_list.

Big-lock approach
-----------------
WHAT IS IT?
Big lock is a coarse-grained lock which protects all three
lists, mentioned above, from racy access.

HOW DOES IT WORK?
At any given point in time, glusterd's thread(s) are in execution
_iff_ there is a preceding, inbound network event. Of course, the
sigwaiter thread and timer thread are exceptions.
A network event is an external trigger to glusterd, via the epoll
thread, in the form of POLLIN and POLLERR.
As long as we take the big-lock at all such entry points and yield
it when we are done, we are guaranteed that all the network events,
accessing the global lists, are serialised.

This amounts to holding the big lock at
- all the handlers of all the actors in glusterd. (POLLIN)
- all the cbks in glusterd. (POLLIN)
- rpc_notify (DISCONNECT event), if we access/modify
  one of the three lists. (POLLERR)

In the case of synctask'ized volume operations, we must remember that,
if we held the big lock for the entire duration of the handler,
we may block other non-synctask rpc actors from executing.
For eg, volume-start would block in PMAP SIGNIN, if done incorrectly.
To prevent this, we need to yield the big lock, when we yield the
synctask, and reacquire on waking up of the synctask.

Change-Id: Ib929f9905b55fb6c3fc27fefb497a26dba058e4f
BUG: 948686
Signed-off-by: Krishnan Parthasarathi 
Reviewed-on: http://review.gluster.org/4784
Reviewed-by: Jeff Darcy 
Tested-by: Gluster Build System

glusterd: Use client-op-versions during "volume set"

2013-02-06T20:17:49+00:00

The supported op-versions of the client and the name of the requested volume,
are saved during server_getspec(). These are used during the staging of volume
set. If the option being set is not supported by any of the clients which
currently have the volume mounted, then set will fail.

Change-Id: I4e6b60b274d5200508762dc0204cfa848a6c0aa4
BUG: 907311
Signed-off-by: Kaushal M 
Reviewed-on: http://review.gluster.org/4424
Tested-by: Gluster Build System 
Reviewed-by: Amar Tumballi 
Reviewed-by: Anand Avati