summaryrefslogtreecommitdiffstats
path: root/xlators
Commit message (Collapse)AuthorAgeFilesLines
* glusterd: Disable ping-timer between glusterd and brick processVijaikumar M2014-05-193-6/+10
| | | | | | | | | | | | | | | | | | | | | | When there are too many IO happening, brick process epoll thread will be busy and fails to respond to the glusterd pick packet within 30sec. Also epoll thread can be blocked by a big-lock. Solution is to disable ping-timer by default and only enable where ever required Later when the epoll thread model changed and made lighter, we need to revert back this change. http://review.gluster.com/3842 is one such approach. Change-Id: I7f80ad3eb00f7d9c4d4527305932f7cf4920e73f BUG: 1097224 Signed-off-by: Vijaikumar M <vmallika@redhat.com> Reviewed-on: http://review.gluster.org/7753 Reviewed-by: Krishnan Parthasarathi <kparthas@redhat.com> Tested-by: Gluster Build System <jenkins@build.gluster.com> Reviewed-by: Vijay Bellur <vbellur@redhat.com>
* rpcsvc: Validate RPC procedure number before fetchSantosh Kumar Pradhan2014-05-179-21/+22
| | | | | | | | | | | | | | | | | | | | | 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. Special Thanks to: Murray Ketchion, Grant Byers Change-Id: I8b5abd406d47fab8fca65b3beb73cdfe8cd85b72 BUG: 1096020 Signed-off-by: Santosh Kumar Pradhan <spradhan@redhat.com> Reviewed-on: http://review.gluster.org/7726 Tested-by: Gluster Build System <jenkins@build.gluster.com> Reviewed-by: Rajesh Joseph <rjoseph@redhat.com> Reviewed-by: Anand Avati <avati@redhat.com>
* meta: add .file_write method into meta_opsAnand Avati2014-05-174-77/+38
| | | | | | | | | | | | | Add .file_write method in meta_ops and make write support into virtual files (like loglevel, measure_latency) much simpler. Change-Id: I812f782f645605c56f81b158bb0f289b3c8f909a BUG: 1089216 Signed-off-by: Anand Avati <avati@redhat.com> Reviewed-on: http://review.gluster.org/7785 Reviewed-by: Krishnan Parthasarathi <kparthas@redhat.com> Reviewed-by: Vijay Bellur <vbellur@redhat.com> Tested-by: Gluster Build System <jenkins@build.gluster.com>
* NetBSD build fixesEmmanuel Dreyfus2014-05-173-29/+34
| | | | | | | | | | | | | | | | | | | | | | | | | | | | - Shell scripts: == is specific to bash and ksh. Use = instead. - Shell scripts: use sh instead of bash if bash functionnality is not used - Shell scripts: ${var/search/replace} is specific to bash - sed: The -i option is specific to GNU sed. - Makefiles: $< outside of generic rules only work in GNU make. - xdrproc_t() is not universally defined as variadic. Do not specify third argument if it is not used - NetBSD FUSE specific: only include <perfuse.h> in FUSE client code, it harms in other locations - configure: Search for gettext() in libintl as NetBSD stores it there - Like MacOS X, NetBSD has unmount(2) and not umount(2) (un vs u) Some other build issues previously included in this change were removed: - __THROW macro, addressed in http://review.gluster.com/#/c/7757/ - getmntent() compat shared with MacOS X, in http://review.gluster.com/#/c/7722/ This patchset adds warning fixes for mount_glusterfs BUG: 764655 Change-Id: I2f1faf8ff96362d3e2baf237b943df619011f1f4 Signed-off-by: Emmanuel Dreyfus <manu@netbsd.org> Reviewed-on: http://review.gluster.org/7783 Tested-by: Gluster Build System <jenkins@build.gluster.com> Reviewed-by: Harshavardhana <harsha@harshavardhana.net>
* contrib: Cross platform fixes after recent commitsHarshavardhana2014-05-172-10/+10
| | | | | | | | | | | | | | | - provide a getment_r () version which behaves as re-entrant with some caveats for NetBSD/OSX specific. - some apparent warning issues fixed, always use PRI* format specification avoid using %ld i.e not portable Change-Id: Ib3d1a73b426e38b436b356355b97db0104a1a4a5 BUG: 1089172 Signed-off-by: Harshavardhana <harsha@harshavardhana.net> Reviewed-on: http://review.gluster.org/7722 Reviewed-by: Emmanuel Dreyfus <manu@netbsd.org> Tested-by: Gluster Build System <jenkins@build.gluster.com> Reviewed-by: Anand Avati <avati@redhat.com>
* cluster/afr: move messages to new logging frameworkRavishankar N2014-05-1711-21/+93
| | | | | | | | | | | | | | Change important (from a diagnostics point of view) log messages to use the gf_msg() framework. Change-Id: I0a58184bbb78989db149e67f07c140a21c781bc2 BUG: 1075611 Signed-off-by: Ravishankar N <ravishankar@redhat.com> Reviewed-on: http://review.gluster.org/7784 Reviewed-by: Krutika Dhananjay <kdhananj@redhat.com> Tested-by: Gluster Build System <jenkins@build.gluster.com> Reviewed-by: Pranith Kumar Karampuri <pkarampu@redhat.com> Tested-by: Pranith Kumar Karampuri <pkarampu@redhat.com>
* protocol/client,server: Suppress ESTALE logsPranith Kumar K2014-05-162-61/+75
| | | | | | | | | | | Change-Id: I5c5e016138be2417eacf81812313dc854907fa47 BUG: 1095256 Signed-off-by: Pranith Kumar K <pkarampu@redhat.com> Reviewed-on: http://review.gluster.org/7696 Reviewed-by: Atin Mukherjee <amukherj@redhat.com> Tested-by: Gluster Build System <jenkins@build.gluster.com> Reviewed-by: Raghavendra G <rgowdapp@redhat.com> Tested-by: Raghavendra G <rgowdapp@redhat.com>
* glusterd/snapshot: Putting back the missed_snaps_list codeAvra Sengupta2014-05-151-1/+8
| | | | | | | | | | | | | | Setting of missed_snap_count was removed as part of an earlier patch. Putting back the code. Signed-off-by: Avra Sengupta <asengupt@redhat.com> Change-Id: Ib6412d6100145e94d10f6f4a8a1fe4e645c1a69e BUG: 1097725 Reviewed-on: http://review.gluster.org/7764 Tested-by: Gluster Build System <jenkins@build.gluster.com> Reviewed-by: Rajesh Joseph <rjoseph@redhat.com> Reviewed-by: Atin Mukherjee <amukherj@redhat.com> Reviewed-by: Kaushal M <kaushal@redhat.com>
* glusterd: Differentiate snap-volume directories properly.Avra Sengupta2014-05-151-4/+3
| | | | | | | | | | | | | | | | | | If /var/lib/glusterd is hosted on xfs system, the entry->d_type not showing the correct d_type owes to the restore path prematurely exiting. Hence checking entry->d_name to differntiate <snap-name>/info file, missed_snaps_list file and the <snap-name>/geo-replication directory, from the actual volume directories, without impacting the gluster volumes. Change-Id: I9a774a845282fe7cc697e37bbcf7c4545aee7678 BUG: 1094557 Signed-off-by: Avra Sengupta <asengupt@redhat.com> Reviewed-on: http://review.gluster.org/7680 Tested-by: Gluster Build System <jenkins@build.gluster.com> Reviewed-by: Sachin Pandit <spandit@redhat.com> Reviewed-by: Krutika Dhananjay <kdhananj@redhat.com> Reviewed-by: Kaushal M <kaushal@redhat.com>
* cluster/afr: Fix bugs in quorum implementationPranith Kumar K2014-05-148-102/+164
| | | | | | | | | | | | | - Have common place to perform quorum fop wind check - Check if fop succeeded in a way that matches quorum to avoid marking changelog in split-brain. BUG: 1066996 Change-Id: Ibc5b80e01dc206b2abbea2d29e26f3c60ff4f204 Signed-off-by: Pranith Kumar K <pkarampu@redhat.com> Reviewed-on: http://review.gluster.org/7600 Tested-by: Gluster Build System <jenkins@build.gluster.com> Reviewed-by: Ravishankar N <ravishankar@redhat.com>
* glusterd/geo-rep: Allow gverify.sh and S56glusterd-geo-rep-create-post.shAvra Sengupta2014-05-141-16/+68
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | to operate for non-root privileged slave volume Mounting the slave-volume on local node, to perform disk checks in order to allow gverify.sh to operate for non-root privileged slave volume Allowing the hook script S56glusterd-geo-rep-create-post.sh to operate for non-root privileged slave volume Modified peer_add_secret_pub.in to accept username as argument and add the pem keys to the users's_home_dir/.ssh/authorized_keys Wrote set_geo_rep_pem_keys.sh which accepts username as argument and copies the pem keys from the user's home directory to $GLUSTERD_WORKING_DIR/geo-replication/ and then copies the keys to other nodes in the cluster and add them to the respective authorized keys. The script takes as argument the user name and assumes that the user will be present in all the nodes in the cluster. It is not needed for root. To summarize: For a privileged slave user, execute the following on master node as super user: gluster system:: execute gsec_create gluster volume geo-replication <master_vol> [root@]<slave_ip>::<slave_vol> create push_pem For a non-privileged slave user execute the following on master node as super user: gluster system:: execute gsec_create gluster volume geo-replication <master_vol> <slave_user>@<slave_ip>::<slave_vol> create push_pem then on the slave node execute the following as super user: /usr/local/libexec/glusterfs/set_geo_rep_pem_keys.sh <slave_user> BUG: 1077452 Change-Id: I88020968aa5b13a2c2ab86b1d6661b60071f6f5e Signed-off-by: Avra Sengupta <asengupt@redhat.com> Reviewed-on: http://review.gluster.org/7744 Tested-by: Gluster Build System <jenkins@build.gluster.com> Reviewed-by: Venky Shankar <vshankar@redhat.com> Tested-by: Venky Shankar <vshankar@redhat.com>
* gsyncd / geo-rep: Partial support for Non-root geo-replication.Venky Shankar2014-05-145-120/+158
| | | | | | | | | | | | | | | | | | | | | | | | This patch enables geo-replication to be run as an unprivileged user. As of now, this is just the partial support, but is very close to achieve full functionality. Current limitation * Geo-replication executed Gluster CLI commands on the slave via SSH. On a non-root setup, Gluster CLI would run as an unprivileged user, failing to execute the command. As a workaround (for testing), setuid(2) Gluster CLI executable or use the glusterd option to accept commands by unprivileged CLI process. The nature of cli commands are "system::" commands (for key management) and remote volume info fetching. Remote volume info fetching has been modified to use --remote-host gluster cli option rather than ssh and remote cli execution. Change-Id: Ica89e2ba9b7f48fd6e1c876c477d7822dc693617 BUG: 1077452 Signed-off-by: Venky Shankar <vshankar@redhat.com> Reviewed-on: http://review.gluster.org/7658 Tested-by: Gluster Build System <jenkins@build.gluster.com>
* features/changelog : historical journal consumption.Venky Shankar2014-05-149-20/+1032
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Facilitates Glusterfs with the ability to detect file-operations happened in past by scanning the back-end(brick-level) glusterfs journal (changelog). Design: * List of changelogs produces in one perfectly running session are stored in htime file which also holds necessary information about the session start and end time. * Involves fixed sized seeks to identify N'th changelog in the list. * Requires O(log n), (where n is number of changelogs in the list), time to identify the end changelog for the given start-end time interval. Currently the background processing of changelogs is sub optimal. BZ 1097041 tracks the development effort. For complete design, refer the below link: http://lists.nongnu.org/archive/html/gluster-devel/2014-02/msg00206.html Change-Id: I27e49f75e492e843084d0ecaf9130224d08462a0 BUG: 1091961 Signed-off-by: Ajeet Jha <ajha@redhat.com> Signed-off-by: Venky Shankar <vshankar@redhat.com> Signed-off-by: Ajeet Jha <ajha@redhat.com> Reviewed-on: http://review.gluster.org/6930 Reviewed-by: Kotresh HR <khiremat@redhat.com> Tested-by: Gluster Build System <jenkins@build.gluster.com>
* glusterd: Allow setting volume options by default based on op-versionKaushal M2014-05-145-5/+83
| | | | | | | | | | | | | | | | | | A new function glusterd_enable_default_options is introduced, which will set some volume options on a volume based on op-version. This function is called near the end of the volume create and will allow some options to be enabled based on op-version on newly created volumes. This will also be called during volume reset, to reset the options to their default values if they had changed. Change-Id: I91057d9e42409b17a884728b43ae3721328d4831 BUG: 1096616 Signed-off-by: Kaushal M <kaushal@redhat.com> Reviewed-on: http://review.gluster.org/7734 Tested-by: Gluster Build System <jenkins@build.gluster.com> Reviewed-by: Atin Mukherjee <amukherj@redhat.com> Reviewed-by: Krishnan Parthasarathi <kparthas@redhat.com> Tested-by: Krishnan Parthasarathi <kparthas@redhat.com>
* geo-rep/glusterd: Pause and Resume feature for geo-replicationKotresh H R2014-05-131-14/+263
| | | | | | | | | | | | | | | | | This patch introduces pause and resume cli command for geo-replication. Signed-off-by: Kotresh H R <khiremat@redhat.com> Change-Id: I4f5e58e9175fe85077d56088473252391fb57de7 BUG: 1093602 Signed-off-by: Kotresh H R <khiremat@redhat.com> Reviewed-on: http://review.gluster.org/7643 Tested-by: Gluster Build System <jenkins@build.gluster.com> Reviewed-by: Atin Mukherjee <amukherj@redhat.com> Reviewed-by: Avra Sengupta <asengupt@redhat.com> Reviewed-by: Venky Shankar <vshankar@redhat.com> Tested-by: Venky Shankar <vshankar@redhat.com>
* glusterd/snapshot : Quorum check should not be made if weSachin Pandit2014-05-132-129/+163
| | | | | | | | | | | | | | | | | | | | | | | | perform snapshot status command. Problem : Snapshot status command used to fail as it used to hit the quorum check path. Solution : The condition checking where snapname is fetched based on the presence of snap_volume is moved inside create switch case. And also moved the chunk of code which does the actual quorum check to new function to make the code more readable. Change-Id: Idda2d7c576cdfab3a7d087bfa74bfa616372c20e BUG: 1096700 Signed-off-by: Sachin Pandit <spandit@redhat.com> Reviewed-on: http://review.gluster.org/7737 Reviewed-by: Rajesh Joseph <rjoseph@redhat.com> Tested-by: Gluster Build System <jenkins@build.gluster.com> Reviewed-by: Raghavendra Bhat <raghavendra@redhat.com> Reviewed-by: Kaushal M <kaushal@redhat.com>
* glusterd : barrier enable/disable should fail if already enabled/disabledAtin Mukherjee2014-05-121-3/+9
| | | | | | | | | | | | | | | | | In barrier notify function, if we fail to set the barrier option execution goes to default_notify which returns 0 and command returns success. Fix : We need not call the default_notify function when handling GF_EVENT_TRANSLATOR_OP in barrier xlator's notify. Change-Id: Ia2c361b43cca7791c29829d69dcd6fc7923102f6 BUG: 1092841 Signed-off-by: Atin Mukherjee <amukherj@redhat.com> Reviewed-on: http://review.gluster.org/7609 Reviewed-by: Krishnan Parthasarathi <kparthas@redhat.com> Tested-by: Gluster Build System <jenkins@build.gluster.com> Reviewed-by: Sachin Pandit <spandit@redhat.com> Reviewed-by: Kaushal M <kaushal@redhat.com>
* glusterd: On gaining quorum spawn_daemons in new threadKaushal M2014-05-124-28/+44
| | | | | | | | | | | | | | | | | | | | | | | | | | | During startup, if a glusterd has peers, it waits till quorum is obtained to spawn bricks and other services. If peers are not present, the daemons are started during glusterd' startup itself. The spawning of daemons as a quorum action was done without using a seperate thread, unlike the spawn on startup. Since, quotad was launched using the blocking runner_run api, this leads to the thread being blocked. The calling thread is almost always the epoll thread and this leads to a deadlock. The runner_run call blocks the epoll thread waiting for quotad to start, as a result glusterd cannot serve any requests. But the startup of quotad is blocked as it cannot fetch the volfile from glusterd. The fix for this is to launch the spawn daemons task in a seperate thread. This will free up the epoll thread and prevents the above deadlock from happening. Change-Id: Ife47b3591223cdfdfb2b4ea8dcd73e63f18e8749 BUG: 1095585 Signed-off-by: Kaushal M <kaushal@redhat.com> Reviewed-on: http://review.gluster.org/7703 Reviewed-by: Krishnan Parthasarathi <kparthas@redhat.com> Tested-by: Gluster Build System <jenkins@build.gluster.com> Tested-by: Krishnan Parthasarathi <kparthas@redhat.com>
* libgfapi: Added support to fetch volume info from glusterd and store in ↵Soumya Koduri2014-05-111-0/+137
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | glfs object. 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. Change-Id: I303f76d7177d32d25bdb301b1dbcf5cd73f42807 BUG: 1090363 Signed-off-by: Soumya Koduri <skoduri@redhat.com> Reviewed-on: http://review.gluster.org/7218 Reviewed-by: Anand Avati <avati@redhat.com> Reviewed-by: Harshavardhana <harsha@harshavardhana.net> Tested-by: Gluster Build System <jenkins@build.gluster.com> Reviewed-by: Vijay Bellur <vbellur@redhat.com>
* rpc: implement server.manage-gids for group resolving on the bricksNiels de Vos2014-05-097-1/+207
| | | | | | | | | | | | | | | | | | | | | | | | | The new volume option 'server.manage-gids' can be enabled in environments where a user belongs to more than the current absolute maximum of 93 groups. This option triggers the following behavior: 1. The AUTH_GLUSTERFS structure sent by GlusterFS clients (fuse, nfs or libgfapi) will contain only one (1) auxiliary group, instead of a full list. This reduces network usage and prevents problems in encoding the AUTH_GLUSTERFS structure which should fit in 400 bytes. 2. The single group in the RPC Calls received by the server is replaced by resolving the groups server-side. Permission checks and similar in lower xlators are applied against the full list of groups where the user belongs to, and not the single auxiliary group that the client sent. Change-Id: I9e540de13e3022f8b63ff893ecba511129a47b91 BUG: 1053579 Signed-off-by: Niels de Vos <ndevos@redhat.com> Reviewed-on: http://review.gluster.org/7501 Tested-by: Gluster Build System <jenkins@build.gluster.com> Reviewed-by: Santosh Pradhan <spradhan@redhat.com> Reviewed-by: Harshavardhana <harsha@harshavardhana.net> Reviewed-by: Anand Avati <avati@redhat.com>
* mgmt/glusterd: delete oldest snapshot upon exceeding soft-limitRaghavendra Bhat2014-05-082-0/+106
| | | | | | | | | | Change-Id: I2d6ebae3ced1910f2dee43eeb9fc430e9f31073f BUG: 1061685 Signed-off-by: Raghavendra Bhat <raghavendra@redhat.com> Reviewed-on: http://review.gluster.org/7587 Reviewed-by: Rajesh Joseph <rjoseph@redhat.com> Reviewed-by: Krishnan Parthasarathi <kparthas@redhat.com> Tested-by: Krishnan Parthasarathi <kparthas@redhat.com>
* glusterd/snapshot: volume gets deleted if restore failsRajesh Joseph2014-05-081-0/+8
| | | | | | | | | | | | | | | If the restore command fails in pre-validate phase then main volume gets deleted. Fix: Perform cleanup only when pre-validate passes. Change-Id: I7128c8582c3dd166a5683babb7e136ad0b56f0ac BUG: 1061685 Signed-off-by: Rajesh Joseph <rjoseph@redhat.com> Reviewed-on: http://review.gluster.org/7665 Tested-by: Gluster Build System <jenkins@build.gluster.com> Reviewed-by: Krishnan Parthasarathi <kparthas@redhat.com> Tested-by: Krishnan Parthasarathi <kparthas@redhat.com>
* glusterd/snapshot: Don't release big_lock before completing snapshot creationVijaikumar M2014-05-083-5/+8
| | | | | | | | | | | | | | | | | | | | | | | | | | Releasing the big-lock can cause problem like deadlock or memory corruption. Same happened with bug 1091926 where glusterd on node-2 entered a commit phase and released a big-lock. The originator node received timeout for the commit phase and triggered a post-validate cleanup to the node-2. Now node-2 continued to work with the object that are alreday cleaned-up and resulted in a crash. Solution is to not to release big-lock in the commit phase of snapshot creation. Change-Id: I571194fdb0b0ecc91bd13f2a9fc92fe4338d14dc BUG: 1091926 Signed-off-by: Vijaikumar M <vmallika@redhat.com> Reviewed-on: http://review.gluster.org/7579 Tested-by: Gluster Build System <jenkins@build.gluster.com> Reviewed-by: Rajesh Joseph <rjoseph@redhat.com> Reviewed-by: Krishnan Parthasarathi <kparthas@redhat.com> Tested-by: Krishnan Parthasarathi <kparthas@redhat.com>
* glusterd/snapshot: Execute lvm snapshots in parallelVijaikumar M2014-05-085-73/+248
| | | | | | | | | | | | | | | | | Back-end LVM Snapshot is executed parallely as synop task This helps is gaining performance when there are more bricks in a node. This patch also removes unwanted logs printed in snapshot cleanup Change-Id: I3174cb4547ebb670eca37a98eb9d75ecb0672a90 BUG: 1061685 Signed-off-by: Vijaikumar M <vmallika@redhat.com> Reviewed-on: http://review.gluster.org/7461 Tested-by: Gluster Build System <jenkins@build.gluster.com> Reviewed-by: Rajesh Joseph <rjoseph@redhat.com> Reviewed-by: Krishnan Parthasarathi <kparthas@redhat.com> Tested-by: Krishnan Parthasarathi <kparthas@redhat.com>
* glusterd/snapshot: Add brick-count suffix for the LVM snapshotVijaikumar M2014-05-085-278/+245
| | | | | | | | | | | | | | | | When there are more than one brick created from the same LVM volume group, there will be a conflict with the LVM snapshot name we use. Solution is to add a brick-count suffix to the LVM snapshot name Change-Id: I7258e69fe0b50e86b81c66ab1db523ab3c7cbae0 BUG: 1091934 Signed-off-by: Vijaikumar M <vmallika@redhat.com> Reviewed-on: http://review.gluster.org/7581 Tested-by: Gluster Build System <jenkins@build.gluster.com> Reviewed-by: Rajesh Joseph <rjoseph@redhat.com> Reviewed-by: Krishnan Parthasarathi <kparthas@redhat.com> Tested-by: Krishnan Parthasarathi <kparthas@redhat.com>
* cluster/afr: Remove stale index in self-heal codepathPranith Kumar K2014-05-082-5/+7
| | | | | | | | | Change-Id: I635fc0fa955b33590f1c5b4dfec22d591ea8575c BUG: 1032894 Signed-off-by: Pranith Kumar K <pkarampu@redhat.com> Reviewed-on: http://review.gluster.org/6592 Tested-by: Gluster Build System <jenkins@build.gluster.com> Reviewed-by: Anand Avati <avati@redhat.com>
* libgfchangelog: memory accounting intializationVenky Shankar2014-05-081-0/+3
| | | | | | | | | | | Needed as memory accounting turned ON by default now. Without this, GF_{CALLOC,MALLOC} calls assert. Change-Id: Ie54a6efb61871e70d4f7c796d9c86a867051e2e0 BUG: 1094708 Signed-off-by: Venky Shankar <vshankar@redhat.com> Reviewed-on: http://review.gluster.org/7698 Reviewed-by: Kotresh HR <khiremat@redhat.com>
* mgmt/glusterd: Prevent spurious brick restartsPranith Kumar K2014-05-083-23/+44
| | | | | | | | | Change-Id: I7ee5d18b926d6c31e3e4ea2f5fbe9050c8e1dee8 BUG: 959986 Signed-off-by: Pranith Kumar K <pkarampu@redhat.com> Reviewed-on: http://review.gluster.org/4954 Tested-by: Gluster Build System <jenkins@build.gluster.com> Reviewed-by: Kaushal M <kaushal@redhat.com>
* mgmt/glusterd: Perform Pending quorum actions after OpPranith Kumar K2014-05-081-0/+7
| | | | | | | | | Change-Id: I2bb67b5fb4a6f6dac892ef3206e7a79706018a6e BUG: 959986 Signed-off-by: Pranith Kumar K <pkarampu@redhat.com> Reviewed-on: http://review.gluster.org/4955 Tested-by: Gluster Build System <jenkins@build.gluster.com> Reviewed-by: Kaushal M <kaushal@redhat.com>
* glusterd: Use a calloc-ed copy of txn_id for glusterd_do_replace_brickAvra Sengupta2014-05-081-3/+8
| | | | | | | | | | | | | | | As glusterd_do_replace_brick() is spawned through gf_timer_call_after(), by the time it's called the event is freed, and the txn_id is lost. Hence using a calloc-ed copy, which will be freed as a part of rb_ctx dict. Change-Id: I3e309fe1a7ba96ad1d1ce01f4d2aa18178f59244 BUG: 1095097 Signed-off-by: Avra Sengupta <asengupt@redhat.com> Reviewed-on: http://review.gluster.org/7686 Reviewed-by: Krishnan Parthasarathi <kparthas@redhat.com> Tested-by: Gluster Build System <jenkins@build.gluster.com> Reviewed-by: Kaushal M <kaushal@redhat.com>
* mgmt/glusterd: quorum check before taking the snapshotRaghavendra Bhat2014-05-075-23/+667
| | | | | | | | | | | | | | | | | | | | without force option: quorum fails if glusterds are not in quorum. If glusterd are in quorum, then volume quorum (i.e quorum of the bricks) is checked. volume quorum fails even if one of the bricks are down. with force option: even though the glusterds are not in quorum, and some bricks are down, the quorum check of the volume (i.e bricks) is done and if the volume quorum is met, snapshot is taken. Change-Id: I06971e45d5cf09880032ef038bfe011e6c244268 BUG: 1061685 Signed-off-by: Raghavendra Bhat <raghavendra@redhat.com> Reviewed-on: http://review.gluster.org/7463 Tested-by: Gluster Build System <jenkins@build.gluster.com> Reviewed-by: Rajesh Joseph <rjoseph@redhat.com> Reviewed-by: Kaushal M <kaushal@redhat.com>
* glusterd: NFS server wrongly started with `DEBUG` log-levelHarshavardhana2014-05-071-2/+1
| | | | | | | | | | | | | | Disable DEBUG Change-Id: I011231ba3df4a42f892f1305867bfc74bb101269 BUG: 1089172 Signed-off-by: Harshavardhana <harsha@harshavardhana.net> Reviewed-on: http://review.gluster.org/7654 Reviewed-by: Niels de Vos <ndevos@redhat.com> Reviewed-by: Pranith Kumar Karampuri <pkarampu@redhat.com> Tested-by: Gluster Build System <jenkins@build.gluster.com> Reviewed-by: Krishnan Parthasarathi <kparthas@redhat.com> Tested-by: Krishnan Parthasarathi <kparthas@redhat.com>
* glusterd : Port glusterd sync log messages to gf_msg APIAtin Mukherjee2014-05-064-33/+67
| | | | | | | | | Change-Id: Ic3ed2c96d8fc3a15fedaa80517a2c79c0c858963 BUG: 1075611 Signed-off-by: Atin Mukherjee <amukherj@redhat.com> Reviewed-on: http://review.gluster.org/7652 Tested-by: Gluster Build System <jenkins@build.gluster.com> Reviewed-by: Kaushal M <kaushal@redhat.com>
* glusterd: port network failure log messages to gf_msg APIKrishnan Parthasarathi2014-05-063-11/+33
| | | | | | | | | Change-Id: I23df6d179e9d66a71721e9844a34c5b96586f90f BUG: 1075611 Signed-off-by: Krishnan Parthasarathi <kparthas@redhat.com> Reviewed-on: http://review.gluster.org/7462 Tested-by: Gluster Build System <jenkins@build.gluster.com> Reviewed-by: Kaushal M <kaushal@redhat.com>
* glusterd: Port server quorum messages to the gf_msg APIKaushal M2014-05-065-6/+84
| | | | | | | | | | Change-Id: I84716cc07f3cbd8c1b2825a5676d6693fed6fade BUG: 1075611 Signed-off-by: Kaushal M <kaushal@redhat.com> Reviewed-on: http://review.gluster.org/7578 Tested-by: Gluster Build System <jenkins@build.gluster.com> Reviewed-by: Krishnan Parthasarathi <kparthas@redhat.com> Tested-by: Krishnan Parthasarathi <kparthas@redhat.com>
* glusterd: Allow bumping up the cluster op-versionKaushal M2014-05-062-0/+77
| | | | | | | | | | | | | | | | | | | | | | | | | This patch allows a user to bump up the cluster op-version by doing # gluster volume set all cluster.op-version <OP-VERSION> The op-version will be bumped only if - all the peers in the cluster support it, and - the new op-version is greater than the current cluster op-version This set operation will not do any other change other than changing and saving the cluster op-version in the glusterd.info file. It will NOT, - change any existing volume - add the option to the global options list - fix the cluster op-version to the given version, it can be bumped up by other volume set commands. Change-Id: I084b4fcc45e79dc2ca7b7680d7bb371bb175af39 BUG: 1092592 Signed-off-by: Kaushal M <kaushal@redhat.com> Reviewed-on: http://review.gluster.org/7603 Reviewed-by: Atin Mukherjee <amukherj@redhat.com> Tested-by: Gluster Build System <jenkins@build.gluster.com> Reviewed-by: Krishnan Parthasarathi <kparthas@redhat.com> Tested-by: Krishnan Parthasarathi <kparthas@redhat.com>
* glusterd/snashot: Perform missed snap createsAvra Sengupta2014-05-064-145/+589
| | | | | | | | | | | | | | | | | | | | | | | | | When a brick is started, and the glusterfsd process requests for volfile, the brick_name is sent in the req dict. In glusterd, after fetching the spec the brick_name is looked up in the missed_snap_list, and any missing snap creates on the same brick are performed. After this, the glusterd responds back with the specfile. Also collate brick data from the node's hosting the bricks during restore. In case the data is absent, the local node's data is used. This is needed to ensure that, during a restore we collect the information created when a missed snap create is performed. Change-Id: I47cefdeba96f2702be810965734cf0fac61d3d2d BUG: 1061685 Signed-off-by: Avra Sengupta <asengupt@redhat.com> Reviewed-on: http://review.gluster.org/7551 Reviewed-by: Santosh Pradhan <spradhan@redhat.com> Reviewed-by: Rajesh Joseph <rjoseph@redhat.com> Tested-by: Gluster Build System <jenkins@build.gluster.com> Reviewed-by: Krishnan Parthasarathi <kparthas@redhat.com> Tested-by: Krishnan Parthasarathi <kparthas@redhat.com>
* glusterd: Fetch brick mount_dirs during brick create.Avra Sengupta2014-05-0611-117/+461
| | | | | | | | | | | | | | | | | Fetch the mount directory path for a brick, during volume create, add-brick, and replace-brick. When a snap-create is missed, use this mount directory information to create the brick path for the missed snap brick. Change-Id: Iad3eec96a32cf340f26bdf3f28e2f529e4b77e31 BUG: 1061685 Signed-off-by: Avra Sengupta <asengupt@redhat.com> Reviewed-on: http://review.gluster.org/7550 Reviewed-by: Rajesh Joseph <rjoseph@redhat.com> Tested-by: Gluster Build System <jenkins@build.gluster.com> Reviewed-by: Krishnan Parthasarathi <kparthas@redhat.com> Tested-by: Krishnan Parthasarathi <kparthas@redhat.com>
* meta: print in json for stack/frames, cmdline and versionHarshavardhana2014-05-053-71/+99
| | | | | | | | | | | | | - Follow formatting rules based on RFC4627 - http://www.ietf.org/rfc/rfc4627.txt - Add checks for json in regression test meta.t Change-Id: I480d32ce042b202d3ed8939623c629a03b458551 BUG: 1089216 Signed-off-by: Harshavardhana <harsha@harshavardhana.net> Reviewed-on: http://review.gluster.org/7653 Tested-by: Gluster Build System <jenkins@build.gluster.com> Reviewed-by: Anand Avati <avati@redhat.com>
* meta: export process state dumpAnand Avati2014-05-0530-22/+419
| | | | | | | | | | | | Various parts of process state dump are exposed through most appropriate virual files/dirs Change-Id: Iff30f8ea76552f2f6d9c98def688c5ec4bc6f2af BUG: 1089216 Signed-off-by: Anand Avati <avati@redhat.com> Reviewed-on: http://review.gluster.org/7542 Tested-by: Gluster Build System <jenkins@build.gluster.com> Reviewed-by: Harshavardhana <harsha@harshavardhana.net>
* meta: (re-)Implement Meta translatorAnand Avati2014-05-0535-1742/+2658
| | | | | | | | | | | | | | | | | The meta translator exposes details about glusterfs itself in the form of a virtual namespace. Loading the translator on the client side creates the meta virtual view under $mntpoint/.meta by default. The directory is not listed (even with ls -a) and can be accessed by doing a "cd /mnt/.meta" Change-Id: I5ffdf39203841a9562a8280a1f79dc76d4dded5d BUG: 1089216 Signed-off-by: Anand Avati <avati@redhat.com> Reviewed-on: http://review.gluster.org/7509 Tested-by: Gluster Build System <jenkins@build.gluster.com> Reviewed-by: Harshavardhana <harsha@harshavardhana.net>
* mgmt/gluster: Use fsync instead of O_SYNCPranith Kumar K2014-05-054-41/+21
| | | | | | | | | | | | | | | | | Glusterd uses O_SYNC to write to temp file then performs renames to the actual file and performs fsync on parent directory. Until this rename happens syncing writes to the file can be deferred. In this patch O_SYNC open of temp file is removed and fsync of the fd before rename is done. Change-Id: Ie7da161b0daec845c7dcfab4154cc45c2f49d825 BUG: 908277 Signed-off-by: Pranith Kumar K <pkarampu@redhat.com> Reviewed-on: http://review.gluster.org/7370 Reviewed-by: Krishnan Parthasarathi <kparthas@redhat.com> Reviewed-by: Raghavendra Bhat <raghavendra@redhat.com> Tested-by: Gluster Build System <jenkins@build.gluster.com> Reviewed-by: Anand Avati <avati@redhat.com>
* changelog: in C99 `inline` shouldn't be part of function definitionHarshavardhana2014-05-032-5/+6
| | | | | | | | | | | | | | | | | - C99 standard dictates that `inline` should be used only to hint function implementation not as a definition - http://clang.llvm.org/compatibility.html#inline This also leads to run time issues as 'undefined symbols' - Remember to use adding explicit headers when using an external function Change-Id: I66a9030da7ac38ded6e0b6a977182be977382fa2 BUG: 1089172 Signed-off-by: Harshavardhana <harsha@harshavardhana.net> Reviewed-on: http://review.gluster.org/7655 Reviewed-by: Anand Avati <avati@redhat.com> Tested-by: Anand Avati <avati@redhat.com>
* glusterd/snapshot: umount2 on OSX/NetBSD is unmountHarshavardhana2014-05-031-9/+21
| | | | | | | | | | Change-Id: I8de4d47bb2a54b915243ea029cce2585fba34876 BUG: 1089172 Signed-off-by: Harshavardhana <harsha@harshavardhana.net> Reviewed-on: http://review.gluster.org/7651 Reviewed-by: Justin Clift <justin@gluster.org> Tested-by: Justin Clift <justin@gluster.org> Reviewed-by: Anand Avati <avati@redhat.com>
* glusterd: Allow symlink parent for snap_mount_folderHarshavardhana2014-05-031-1/+1
| | | | | | | | | | | | If '/var' is a symlink which is on OSX, 'glusterd' initialization fails which is not necessary fix it. Change-Id: I83adc16cfc0e0deaa18acf74ba99299ba4a21d60 BUG: 1061685 Signed-off-by: Harshavardhana <harsha@harshavardhana.net> Reviewed-on: http://review.gluster.org/7558 Tested-by: Gluster Build System <jenkins@build.gluster.com> Reviewed-by: Anand Avati <avati@redhat.com>
* protocol/server: Validate outstanding-rpc-limitSantosh Kumar Pradhan2014-05-032-6/+9
| | | | | | | | | | | | | | | server.outstanding-rpc-limit by passes validation enforcement of gluster as protocol/server ignores any keys starting with rpc (rpc* wildcard). Which also causes a side effect, i.e. help text gets displayed as NULL. Change-Id: I925e028529d671e1e29a0c7ff7e625fdb49a1be4 BUG: 1091364 Signed-off-by: Santosh Kumar Pradhan <spradhan@redhat.com> Reviewed-on: http://review.gluster.org/7562 Reviewed-by: Poornima G <pgurusid@redhat.com> Reviewed-by: Krishnan Parthasarathi <kparthas@redhat.com> Tested-by: Gluster Build System <jenkins@build.gluster.com> Reviewed-by: Anand Avati <avati@redhat.com>
* glusterd : Volname, brickpath & volfpath length validationAtin Mukherjee2014-05-036-21/+61
| | | | | | | | | | | | | | | | | | | | | | | | While creating a volume and adding a brick validation for _POSIX_PATH_MAX is done on absolute pathname instead of relative pathname due to which a brickpath having less than _POSIX_PATH_MAX may also fail the validation if the directory length is greater than (_POSIX_PATH_MAX -strlen(brickpath/volume name). Also this fix addresses one cli response message correction which says the volume file is too long instead of brick path is too long (when brickpath length validation doesn't fail and vol file length validation fails.) It is also important to note that with the current design of volfile naming, it can not be guranteed that volname and brickpath can have max of _POSIX_PATH_MAX characters. Change-Id: I1283d1f9dea96ae797620002c8723719f26a866d BUG: 1085330 Signed-off-by: Atin Mukherjee <amukherj@redhat.com> Reviewed-on: http://review.gluster.org/7420 Reviewed-by: Niels de Vos <ndevos@redhat.com> Reviewed-by: Krishnan Parthasarathi <kparthas@redhat.com> Tested-by: Gluster Build System <jenkins@build.gluster.com> Reviewed-by: Vijay Bellur <vbellur@redhat.com>
* mgmt/glusterd: handle postvalidate carefully when prevalidate failsRaghavendra Bhat2014-05-033-24/+59
| | | | | | | | | | | | | | * Also changed the order of peers retrieval and snapshot retrieval upon glusterd start, so that the snapshot bricks can be properly resolved while cleaning up the snapshots. Change-Id: I120704e4412a9cadb8d90a9b7969f2b4a1196bc5 BUG: 1061685 Signed-off-by: Raghavendra Bhat <raghavendra@redhat.com> Reviewed-on: http://review.gluster.org/7494 Reviewed-by: Kaushal M <kaushal@redhat.com> Tested-by: Gluster Build System <jenkins@build.gluster.com> Reviewed-by: Vijay Bellur <vbellur@redhat.com>
* barrier : fsync call over NFS should be barriered when barrier is enabledAtin Mukherjee2014-05-031-1/+1
| | | | | | | | | | | | | | | | | | | | | | | | | barrier_fsync was not getting hit when fsync call is issued from the application over NFS mount. Investigation reveals that NFS doesn't send an explicit fsync call to the glusterfs server, however NFS converts it with a stable write with O_DSYNC flag, so for a fsync call over NFS mount point, barrier feature always needs to check for writev and its correspoding flags. So to meet all the coditions of write with O_SYNC and fsync getting blocked when barrier is enabled the check has been modified with following: if (!((flags | fd->flags) & (O_SYNC | O_DSYNC))) Point to be noted here is we need to check both fd's flag and the explicit flag as for NFS write fd->flags is 0. Bug ID : 1091902 Change-Id: Ifd5377fca2cc93bb72158dd525b2aab8b4164ca8 Signed-off-by: Atin Mukherjee <amukherj@redhat.com> Reviewed-on: http://review.gluster.org/7633 Reviewed-by: Santosh Pradhan <spradhan@redhat.com> Reviewed-by: Krishnan Parthasarathi <kparthas@redhat.com> Tested-by: Gluster Build System <jenkins@build.gluster.com> Reviewed-by: Vijay Bellur <vbellur@redhat.com>
* barrier : barrier O_SYNC write incorrect flag checkAtin Mukherjee2014-05-031-1/+1
| | | | | | | | | | | | | | | | | | | | | barrier_writev function was doing the following check to determine whether its a O_SYNC write or not: if (!(flags & O_SYNC)) The problem here is this flag is not fd's flag and gfapi write does not copy open call fd's flag into write flag because of which O_SYNC writes were not getting barriered even if barrier was enabled. The check has been modified as: if (!(fd->flags & (O_SYNC | O_DSYNC))) Change-Id: I07b23852d150b81c7317100ca6d22d082ad897cd BUG: 1090488 Signed-off-by: Atin Mukherjee <amukherj@redhat.com> Reviewed-on: http://review.gluster.org/7549 Reviewed-by: Varun Shastry <vshastry@redhat.com> Reviewed-by: Santosh Pradhan <spradhan@redhat.com> Tested-by: Gluster Build System <jenkins@build.gluster.com> Reviewed-by: Vijay Bellur <vbellur@redhat.com>