openkylin
/
linux
mirror of https://gitee.com/openkylin/linux.git
9
0
Fork 0
Code Issues Projects Releases Wiki Activity
linux/fs/ceph/mds_client.h

407 lines
12 KiB
C
Raw Normal View History

ceph: MDS client The MDS (metadata server) client is responsible for submitting requests to the MDS cluster and parsing the response. We decide which MDS to submit each request to based on cached information about the current partition of the directory hierarchy across the cluster. A stateful session is opened with each MDS before we submit requests to it, and a mutex is used to control the ordering of messages within each session. An MDS request may generate two responses. The first indicates the operation was a success and returns any result. A second reply is sent when the operation commits to disk. Note that locking on the MDS ensures that the results of updates are visible only to the updating client before the operation commits. Requests are linked to the containing directory so that an fsync will wait for them to commit. If an MDS fails and/or recovers, we resubmit requests as needed. We also reconnect existing capabilities to a recovering MDS to reestablish that shared session state. Old dentry leases are invalidated. Signed-off-by: Sage Weil <sage@newdream.net>
2009-10-07 02:31:09 +08:00
#ifndef _FS_CEPH_MDS_CLIENT_H
#define _FS_CEPH_MDS_CLIENT_H
#include <linux/completion.h>
ceph: use kref for struct ceph_mds_request Signed-off-by: Sage Weil <sage@newdream.net>
2009-12-08 04:31:09 +08:00
#include <linux/kref.h>
ceph: MDS client The MDS (metadata server) client is responsible for submitting requests to the MDS cluster and parsing the response. We decide which MDS to submit each request to based on cached information about the current partition of the directory hierarchy across the cluster. A stateful session is opened with each MDS before we submit requests to it, and a mutex is used to control the ordering of messages within each session. An MDS request may generate two responses. The first indicates the operation was a success and returns any result. A second reply is sent when the operation commits to disk. Note that locking on the MDS ensures that the results of updates are visible only to the updating client before the operation commits. Requests are linked to the containing directory so that an fsync will wait for them to commit. If an MDS fails and/or recovers, we resubmit requests as needed. We also reconnect existing capabilities to a recovering MDS to reestablish that shared session state. Old dentry leases are invalidated. Signed-off-by: Sage Weil <sage@newdream.net>
2009-10-07 02:31:09 +08:00
#include <linux/list.h>
#include <linux/mutex.h>
ceph: use rbtree for mds requests The rbtree is a more appropriate data structure than a radix_tree. It avoids extra memory usage and simplifies the code. It also fixes a bug where the debugfs 'mdsc' file wasn't including the most recent mds request. Signed-off-by: Sage Weil <sage@newdream.net>
2010-02-16 04:08:46 +08:00
#include <linux/rbtree.h>
ceph: MDS client The MDS (metadata server) client is responsible for submitting requests to the MDS cluster and parsing the response. We decide which MDS to submit each request to based on cached information about the current partition of the directory hierarchy across the cluster. A stateful session is opened with each MDS before we submit requests to it, and a mutex is used to control the ordering of messages within each session. An MDS request may generate two responses. The first indicates the operation was a success and returns any result. A second reply is sent when the operation commits to disk. Note that locking on the MDS ensures that the results of updates are visible only to the updating client before the operation commits. Requests are linked to the containing directory so that an fsync will wait for them to commit. If an MDS fails and/or recovers, we resubmit requests as needed. We also reconnect existing capabilities to a recovering MDS to reestablish that shared session state. Old dentry leases are invalidated. Signed-off-by: Sage Weil <sage@newdream.net>
2009-10-07 02:31:09 +08:00
#include <linux/spinlock.h>
ceph: factor out libceph from Ceph file system This factors out protocol and low-level storage parts of ceph into a separate libceph module living in net/ceph and include/linux/ceph. This is mostly a matter of moving files around. However, a few key pieces of the interface change as well: - ceph_client becomes ceph_fs_client and ceph_client, where the latter captures the mon and osd clients, and the fs_client gets the mds client and file system specific pieces. - Mount option parsing and debugfs setup is correspondingly broken into two pieces. - The mon client gets a generic handler callback for otherwise unknown messages (mds map, in this case). - The basic supported/required feature bits can be expanded (and are by ceph_fs_client). No functional change, aside from some subtle error handling cases that got cleaned up in the refactoring process. Signed-off-by: Sage Weil <sage@newdream.net>
2010-04-07 06:14:15 +08:00
#include <linux/ceph/types.h>
#include <linux/ceph/messenger.h>
#include <linux/ceph/mdsmap.h>
ceph: define ceph_auth_handshake type The definitions for the ceph_mds_session and ceph_osd both contain five fields related only to "authorizers." Encapsulate those fields into their own struct type, allowing for better isolation in some upcoming patches. Fix the #includes in "linux/ceph/osd_client.h" to lay out their more complete canonical path. Signed-off-by: Alex Elder <elder@inktank.com> Reviewed-by: Sage Weil <sage@inktank.com>
2012-05-17 04:16:38 +08:00
#include <linux/ceph/auth.h>
ceph: MDS client The MDS (metadata server) client is responsible for submitting requests to the MDS cluster and parsing the response. We decide which MDS to submit each request to based on cached information about the current partition of the directory hierarchy across the cluster. A stateful session is opened with each MDS before we submit requests to it, and a mutex is used to control the ordering of messages within each session. An MDS request may generate two responses. The first indicates the operation was a success and returns any result. A second reply is sent when the operation commits to disk. Note that locking on the MDS ensures that the results of updates are visible only to the updating client before the operation commits. Requests are linked to the containing directory so that an fsync will wait for them to commit. If an MDS fails and/or recovers, we resubmit requests as needed. We also reconnect existing capabilities to a recovering MDS to reestablish that shared session state. Old dentry leases are invalidated. Signed-off-by: Sage Weil <sage@newdream.net>
2009-10-07 02:31:09 +08:00
/*
* Some lock dependencies:
*
* session->s_mutex
* mdsc->mutex
*
* mdsc->snap_rwsem
*
ceph: use i_ceph_lock instead of i_lock We have been using i_lock to protect all kinds of data structures in the ceph_inode_info struct, including lists of inodes that we need to iterate over while avoiding races with inode destruction. That requires grabbing a reference to the inode with the list lock protected, but igrab() now takes i_lock to check the inode flags. Changing the list lock ordering would be a painful process. However, using a ceph-specific i_ceph_lock in the ceph inode instead of i_lock is a simple mechanical change and avoids the ordering constraints imposed by igrab(). Reported-by: Amon Ott <a.ott@m-privacy.de> Signed-off-by: Sage Weil <sage@newdream.net>
2011-12-01 01:47:09 +08:00
* ci->i_ceph_lock
ceph: MDS client The MDS (metadata server) client is responsible for submitting requests to the MDS cluster and parsing the response. We decide which MDS to submit each request to based on cached information about the current partition of the directory hierarchy across the cluster. A stateful session is opened with each MDS before we submit requests to it, and a mutex is used to control the ordering of messages within each session. An MDS request may generate two responses. The first indicates the operation was a success and returns any result. A second reply is sent when the operation commits to disk. Note that locking on the MDS ensures that the results of updates are visible only to the updating client before the operation commits. Requests are linked to the containing directory so that an fsync will wait for them to commit. If an MDS fails and/or recovers, we resubmit requests as needed. We also reconnect existing capabilities to a recovering MDS to reestablish that shared session state. Old dentry leases are invalidated. Signed-off-by: Sage Weil <sage@newdream.net>
2009-10-07 02:31:09 +08:00
* mdsc->snap_flush_lock
* mdsc->cap_delay_lock
*
*/
ceph: factor out libceph from Ceph file system This factors out protocol and low-level storage parts of ceph into a separate libceph module living in net/ceph and include/linux/ceph. This is mostly a matter of moving files around. However, a few key pieces of the interface change as well: - ceph_client becomes ceph_fs_client and ceph_client, where the latter captures the mon and osd clients, and the fs_client gets the mds client and file system specific pieces. - Mount option parsing and debugfs setup is correspondingly broken into two pieces. - The mon client gets a generic handler callback for otherwise unknown messages (mds map, in this case). - The basic supported/required feature bits can be expanded (and are by ceph_fs_client). No functional change, aside from some subtle error handling cases that got cleaned up in the refactoring process. Signed-off-by: Sage Weil <sage@newdream.net>
2010-04-07 06:14:15 +08:00
struct ceph_fs_client;
ceph: MDS client The MDS (metadata server) client is responsible for submitting requests to the MDS cluster and parsing the response. We decide which MDS to submit each request to based on cached information about the current partition of the directory hierarchy across the cluster. A stateful session is opened with each MDS before we submit requests to it, and a mutex is used to control the ordering of messages within each session. An MDS request may generate two responses. The first indicates the operation was a success and returns any result. A second reply is sent when the operation commits to disk. Note that locking on the MDS ensures that the results of updates are visible only to the updating client before the operation commits. Requests are linked to the containing directory so that an fsync will wait for them to commit. If an MDS fails and/or recovers, we resubmit requests as needed. We also reconnect existing capabilities to a recovering MDS to reestablish that shared session state. Old dentry leases are invalidated. Signed-off-by: Sage Weil <sage@newdream.net>
2009-10-07 02:31:09 +08:00
struct ceph_cap;
/*
* parsed info about a single inode. pointers are into the encoded
* on-wire structures within the mds reply message payload.
*/
struct ceph_mds_reply_info_in {
struct ceph_mds_reply_inode *in;
ceph: implement DIRLAYOUTHASH feature to get dir layout from MDS This implements the DIRLAYOUTHASH protocol feature, which passes the dir layout over the wire from the MDS. This gives the client knowledge of the correct hash function to use for mapping dentries among dir fragments. Note that if this feature is _not_ present on the client but is on the MDS, the client may misdirect requests. This will result in a forward and degrade performance. It may also result in inaccurate NFS filehandle generation, which will prevent fh resolution when the inode is not present in the client cache and the parent directories have been fragmented. Signed-off-by: Sage Weil <sage@newdream.net>
2010-12-15 09:37:52 +08:00
struct ceph_dir_layout dir_layout;
ceph: MDS client The MDS (metadata server) client is responsible for submitting requests to the MDS cluster and parsing the response. We decide which MDS to submit each request to based on cached information about the current partition of the directory hierarchy across the cluster. A stateful session is opened with each MDS before we submit requests to it, and a mutex is used to control the ordering of messages within each session. An MDS request may generate two responses. The first indicates the operation was a success and returns any result. A second reply is sent when the operation commits to disk. Note that locking on the MDS ensures that the results of updates are visible only to the updating client before the operation commits. Requests are linked to the containing directory so that an fsync will wait for them to commit. If an MDS fails and/or recovers, we resubmit requests as needed. We also reconnect existing capabilities to a recovering MDS to reestablish that shared session state. Old dentry leases are invalidated. Signed-off-by: Sage Weil <sage@newdream.net>
2009-10-07 02:31:09 +08:00
u32 symlink_len;
char *symlink;
u32 xattr_len;
char *xattr_data;
ceph: parse inline data in MClientReply and MClientCaps Signed-off-by: Yan, Zheng <zyan@redhat.com>
2014-11-14 21:29:55 +08:00
u64 inline_version;
u32 inline_len;
char *inline_data;
ceph: MDS client The MDS (metadata server) client is responsible for submitting requests to the MDS cluster and parsing the response. We decide which MDS to submit each request to based on cached information about the current partition of the directory hierarchy across the cluster. A stateful session is opened with each MDS before we submit requests to it, and a mutex is used to control the ordering of messages within each session. An MDS request may generate two responses. The first indicates the operation was a success and returns any result. A second reply is sent when the operation commits to disk. Note that locking on the MDS ensures that the results of updates are visible only to the updating client before the operation commits. Requests are linked to the containing directory so that an fsync will wait for them to commit. If an MDS fails and/or recovers, we resubmit requests as needed. We also reconnect existing capabilities to a recovering MDS to reestablish that shared session state. Old dentry leases are invalidated. Signed-off-by: Sage Weil <sage@newdream.net>
2009-10-07 02:31:09 +08:00
};
/*
ceph: Handle file locks in replies from the MDS. Previously the kernel client incorrectly assumed everything was a directory. Signed-off-by: Herb Shiu <herb_shiu@tcloudcomputing.com> Acked-by: Greg Farnum <gregf@hq.newdream.net> Signed-off-by: Sage Weil <sage@newdream.net>
2010-12-02 06:14:38 +08:00
* parsed info about an mds reply, including information about
* either: 1) the target inode and/or its parent directory and dentry,
* and directory contents (for readdir results), or
* 2) the file range lock info (for fcntl F_GETLK results).
ceph: MDS client The MDS (metadata server) client is responsible for submitting requests to the MDS cluster and parsing the response. We decide which MDS to submit each request to based on cached information about the current partition of the directory hierarchy across the cluster. A stateful session is opened with each MDS before we submit requests to it, and a mutex is used to control the ordering of messages within each session. An MDS request may generate two responses. The first indicates the operation was a success and returns any result. A second reply is sent when the operation commits to disk. Note that locking on the MDS ensures that the results of updates are visible only to the updating client before the operation commits. Requests are linked to the containing directory so that an fsync will wait for them to commit. If an MDS fails and/or recovers, we resubmit requests as needed. We also reconnect existing capabilities to a recovering MDS to reestablish that shared session state. Old dentry leases are invalidated. Signed-off-by: Sage Weil <sage@newdream.net>
2009-10-07 02:31:09 +08:00
*/
struct ceph_mds_reply_info_parsed {
struct ceph_mds_reply_head *head;
ceph: Handle file locks in replies from the MDS. Previously the kernel client incorrectly assumed everything was a directory. Signed-off-by: Herb Shiu <herb_shiu@tcloudcomputing.com> Acked-by: Greg Farnum <gregf@hq.newdream.net> Signed-off-by: Sage Weil <sage@newdream.net>
2010-12-02 06:14:38 +08:00
/* trace */
ceph: MDS client The MDS (metadata server) client is responsible for submitting requests to the MDS cluster and parsing the response. We decide which MDS to submit each request to based on cached information about the current partition of the directory hierarchy across the cluster. A stateful session is opened with each MDS before we submit requests to it, and a mutex is used to control the ordering of messages within each session. An MDS request may generate two responses. The first indicates the operation was a success and returns any result. A second reply is sent when the operation commits to disk. Note that locking on the MDS ensures that the results of updates are visible only to the updating client before the operation commits. Requests are linked to the containing directory so that an fsync will wait for them to commit. If an MDS fails and/or recovers, we resubmit requests as needed. We also reconnect existing capabilities to a recovering MDS to reestablish that shared session state. Old dentry leases are invalidated. Signed-off-by: Sage Weil <sage@newdream.net>
2009-10-07 02:31:09 +08:00
struct ceph_mds_reply_info_in diri, targeti;
struct ceph_mds_reply_dirfrag *dirfrag;
char *dname;
u32 dname_len;
struct ceph_mds_reply_lease *dlease;
ceph: Handle file locks in replies from the MDS. Previously the kernel client incorrectly assumed everything was a directory. Signed-off-by: Herb Shiu <herb_shiu@tcloudcomputing.com> Acked-by: Greg Farnum <gregf@hq.newdream.net> Signed-off-by: Sage Weil <sage@newdream.net>
2010-12-02 06:14:38 +08:00
/* extra */
union {
/* for fcntl F_GETLK results */
struct ceph_filelock *filelock_reply;
/* for readdir results */
struct {
struct ceph_mds_reply_dirfrag *dir_dir;
ceph: preallocate buffer for readdir reply Preallocate buffer for readdir reply. Limit number of entries in readdir reply according to the buffer size. Signed-off-by: Yan, Zheng <zheng.z.yan@intel.com>
2014-03-29 13:41:15 +08:00
size_t dir_buf_size;
ceph: Handle file locks in replies from the MDS. Previously the kernel client incorrectly assumed everything was a directory. Signed-off-by: Herb Shiu <herb_shiu@tcloudcomputing.com> Acked-by: Greg Farnum <gregf@hq.newdream.net> Signed-off-by: Sage Weil <sage@newdream.net>
2010-12-02 06:14:38 +08:00
int dir_nr;
char **dir_dname;
u32 *dir_dname_len;
struct ceph_mds_reply_lease **dir_dlease;
struct ceph_mds_reply_info_in *dir_in;
u8 dir_complete, dir_end;
};
ceph: Check for created flag in response from mds The mds now sends back a created inode if the create request performed the create. If the file already existed, no inode is returned in the reply. This allows ceph to set the created flag in atomic_open so that permissions are properly checked in the case that the file wasn't created by the create call to the mds. To ensure compability with previous kernels, a feature for sending back the inode in the create reply was added, so that the mds will only send back the inode if the client indicates it supports the feature. Signed-off-by: Sam Lang <sam.lang@inktank.com> Reviewed-by: Sage Weil <sage@inktank.com>
2012-12-29 01:56:46 +08:00
/* for create results */
struct {
bool has_create_ino;
u64 ino;
};
ceph: Handle file locks in replies from the MDS. Previously the kernel client incorrectly assumed everything was a directory. Signed-off-by: Herb Shiu <herb_shiu@tcloudcomputing.com> Acked-by: Greg Farnum <gregf@hq.newdream.net> Signed-off-by: Sage Weil <sage@newdream.net>
2010-12-02 06:14:38 +08:00
};
ceph: MDS client The MDS (metadata server) client is responsible for submitting requests to the MDS cluster and parsing the response. We decide which MDS to submit each request to based on cached information about the current partition of the directory hierarchy across the cluster. A stateful session is opened with each MDS before we submit requests to it, and a mutex is used to control the ordering of messages within each session. An MDS request may generate two responses. The first indicates the operation was a success and returns any result. A second reply is sent when the operation commits to disk. Note that locking on the MDS ensures that the results of updates are visible only to the updating client before the operation commits. Requests are linked to the containing directory so that an fsync will wait for them to commit. If an MDS fails and/or recovers, we resubmit requests as needed. We also reconnect existing capabilities to a recovering MDS to reestablish that shared session state. Old dentry leases are invalidated. Signed-off-by: Sage Weil <sage@newdream.net>
2009-10-07 02:31:09 +08:00
/* encoded blob describing snapshot contexts for certain
operations (e.g., open) */
void *snapblob;
int snapblob_len;
};
/*
* cap releases are batched and sent to the MDS en masse.
*/
#define CEPH_CAPS_PER_RELEASE ((PAGE_CACHE_SIZE - \
sizeof(struct ceph_mds_cap_release)) / \
sizeof(struct ceph_mds_cap_item))
/*
* state associated with each MDS<->client session
*/
enum {
CEPH_MDS_SESSION_NEW = 1,
CEPH_MDS_SESSION_OPENING = 2,
CEPH_MDS_SESSION_OPEN = 3,
CEPH_MDS_SESSION_HUNG = 4,
CEPH_MDS_SESSION_CLOSING = 5,
CEPH_MDS_SESSION_RESTARTING = 6,
CEPH_MDS_SESSION_RECONNECTING = 7,
};
struct ceph_mds_session {
struct ceph_mds_client *s_mdsc;
int s_mds;
int s_state;
unsigned long s_ttl; /* time until mds kills us */
u64 s_seq; /* incoming msg seq # */
struct mutex s_mutex; /* serialize session messages */
struct ceph_connection s_con;
ceph: define ceph_auth_handshake type The definitions for the ceph_mds_session and ceph_osd both contain five fields related only to "authorizers." Encapsulate those fields into their own struct type, allowing for better isolation in some upcoming patches. Fix the #includes in "linux/ceph/osd_client.h" to lay out their more complete canonical path. Signed-off-by: Alex Elder <elder@inktank.com> Reviewed-by: Sage Weil <sage@inktank.com>
2012-05-17 04:16:38 +08:00
struct ceph_auth_handshake s_auth;
ceph: negotiate authentication protocol; implement AUTH_NONE protocol When we open a monitor session, we send an initial AUTH message listing the auth protocols we support, our entity name, and (possibly) a previously assigned global_id. The monitor chooses a protocol and responds with an initial message. Initially implement AUTH_NONE, a dummy protocol that provides no security, but works within the new framework. It generates 'authorizers' that are used when connecting to (mds, osd) services that simply state our entity name and global_id. This is a wire protocol change. Signed-off-by: Sage Weil <sage@newdream.net>
2009-11-19 08:19:57 +08:00
ceph: create a new session lock to avoid lock inversion Lockdep was reporting a possible circular lock dependency in dentry_lease_is_valid(). That function needs to sample the session's s_cap_gen and and s_cap_ttl fields coherently, but needs to do so while holding a dentry lock. The s_cap_lock field was being used to protect the two fields, but that can't be taken while holding a lock on a dentry within the session. In most cases, the s_cap_gen and s_cap_ttl fields only get operated on separately. But in three cases they need to be updated together. Implement a new lock to protect the spots updating both fields atomically is required. Signed-off-by: Alex Elder <elder@dreamhost.com> Reviewed-by: Sage Weil <sage@newdream.net>
2012-01-13 09:48:10 +08:00
/* protected by s_gen_ttl_lock */
spinlock_t s_gen_ttl_lock;
ceph: MDS client The MDS (metadata server) client is responsible for submitting requests to the MDS cluster and parsing the response. We decide which MDS to submit each request to based on cached information about the current partition of the directory hierarchy across the cluster. A stateful session is opened with each MDS before we submit requests to it, and a mutex is used to control the ordering of messages within each session. An MDS request may generate two responses. The first indicates the operation was a success and returns any result. A second reply is sent when the operation commits to disk. Note that locking on the MDS ensures that the results of updates are visible only to the updating client before the operation commits. Requests are linked to the containing directory so that an fsync will wait for them to commit. If an MDS fails and/or recovers, we resubmit requests as needed. We also reconnect existing capabilities to a recovering MDS to reestablish that shared session state. Old dentry leases are invalidated. Signed-off-by: Sage Weil <sage@newdream.net>
2009-10-07 02:31:09 +08:00
u32 s_cap_gen; /* inc each time we get mds stale msg */
unsigned long s_cap_ttl; /* when session caps expire */
ceph: create a new session lock to avoid lock inversion Lockdep was reporting a possible circular lock dependency in dentry_lease_is_valid(). That function needs to sample the session's s_cap_gen and and s_cap_ttl fields coherently, but needs to do so while holding a dentry lock. The s_cap_lock field was being used to protect the two fields, but that can't be taken while holding a lock on a dentry within the session. In most cases, the s_cap_gen and s_cap_ttl fields only get operated on separately. But in three cases they need to be updated together. Implement a new lock to protect the spots updating both fields atomically is required. Signed-off-by: Alex Elder <elder@dreamhost.com> Reviewed-by: Sage Weil <sage@newdream.net>
2012-01-13 09:48:10 +08:00
/* protected by s_cap_lock */
spinlock_t s_cap_lock;
ceph: MDS client The MDS (metadata server) client is responsible for submitting requests to the MDS cluster and parsing the response. We decide which MDS to submit each request to based on cached information about the current partition of the directory hierarchy across the cluster. A stateful session is opened with each MDS before we submit requests to it, and a mutex is used to control the ordering of messages within each session. An MDS request may generate two responses. The first indicates the operation was a success and returns any result. A second reply is sent when the operation commits to disk. Note that locking on the MDS ensures that the results of updates are visible only to the updating client before the operation commits. Requests are linked to the containing directory so that an fsync will wait for them to commit. If an MDS fails and/or recovers, we resubmit requests as needed. We also reconnect existing capabilities to a recovering MDS to reestablish that shared session state. Old dentry leases are invalidated. Signed-off-by: Sage Weil <sage@newdream.net>
2009-10-07 02:31:09 +08:00
struct list_head s_caps; /* all caps issued by this session */
int s_nr_caps, s_trim_caps;
int s_num_cap_releases;
ceph: handle race between cap reconnect and cap release When a cap get released while composing the cap reconnect message. We should skip queuing the release message if the cap hasn't been added to the cap reconnect message. Signed-off-by: Yan, Zheng <zheng.z.yan@intel.com> Reviewed-by: Sage Weil <sage@inktank.com>
2013-09-22 11:08:14 +08:00
int s_cap_reconnect;
ceph: handle SESSION_FORCE_RO message mark session as readonly and wake up all cap waiters. Signed-off-by: Yan, Zheng <zyan@redhat.com>
2015-01-05 11:04:04 +08:00
int s_readonly;
ceph: MDS client The MDS (metadata server) client is responsible for submitting requests to the MDS cluster and parsing the response. We decide which MDS to submit each request to based on cached information about the current partition of the directory hierarchy across the cluster. A stateful session is opened with each MDS before we submit requests to it, and a mutex is used to control the ordering of messages within each session. An MDS request may generate two responses. The first indicates the operation was a success and returns any result. A second reply is sent when the operation commits to disk. Note that locking on the MDS ensures that the results of updates are visible only to the updating client before the operation commits. Requests are linked to the containing directory so that an fsync will wait for them to commit. If an MDS fails and/or recovers, we resubmit requests as needed. We also reconnect existing capabilities to a recovering MDS to reestablish that shared session state. Old dentry leases are invalidated. Signed-off-by: Sage Weil <sage@newdream.net>
2009-10-07 02:31:09 +08:00
struct list_head s_cap_releases; /* waiting cap_release messages */
struct list_head s_cap_releases_done; /* ready to send */
ceph: fix iterate_caps removal race We need to be able to iterate over all caps on a session with a possibly slow callback on each cap. To allow this, we used to prevent cap reordering while we were iterating. However, we were not safe from races with removal: removing the 'next' cap would make the next pointer from list_for_each_entry_safe be invalid, and cause a lock up or similar badness. Instead, we keep an iterator pointer in the session pointing to the current cap. As before, we avoid reordering. For removal, if the cap isn't the current cap we are iterating over, we are fine. If it is, we clear cap->ci (to mark the cap as pending removal) but leave it in the session list. In iterate_caps, we can safely finish removal and get the next cap pointer. While we're at it, clean up put_cap to not take a cap reservation context, as it was never used. Signed-off-by: Sage Weil <sage@newdream.net>
2010-02-17 03:39:45 +08:00
struct ceph_cap *s_cap_iterator;
ceph: MDS client The MDS (metadata server) client is responsible for submitting requests to the MDS cluster and parsing the response. We decide which MDS to submit each request to based on cached information about the current partition of the directory hierarchy across the cluster. A stateful session is opened with each MDS before we submit requests to it, and a mutex is used to control the ordering of messages within each session. An MDS request may generate two responses. The first indicates the operation was a success and returns any result. A second reply is sent when the operation commits to disk. Note that locking on the MDS ensures that the results of updates are visible only to the updating client before the operation commits. Requests are linked to the containing directory so that an fsync will wait for them to commit. If an MDS fails and/or recovers, we resubmit requests as needed. We also reconnect existing capabilities to a recovering MDS to reestablish that shared session state. Old dentry leases are invalidated. Signed-off-by: Sage Weil <sage@newdream.net>
2009-10-07 02:31:09 +08:00
/* protected by mutex */
struct list_head s_cap_flushing; /* inodes w/ flushing caps */
struct list_head s_cap_snaps_flushing;
unsigned long s_renew_requested; /* last time we sent a renew req */
u64 s_renew_seq;
atomic_t s_ref;
struct list_head s_waiting; /* waiting requests */
struct list_head s_unsafe; /* unsafe requests */
};
/*
* modes of choosing which MDS to send a request to
*/
enum {
USE_ANY_MDS,
USE_RANDOM_MDS,
USE_AUTH_MDS, /* prefer authoritative mds for this metadata item */
};
struct ceph_mds_request;
struct ceph_mds_client;
/*
* request completion callback
*/
typedef void (*ceph_mds_request_callback_t) (struct ceph_mds_client *mdsc,
struct ceph_mds_request *req);
ceph: fix file lock interruption When a lock operation is interrupted, current code sends a unlock request to MDS to undo the lock operation. This method does not work as expected because the unlock request can drop locks that have already been acquired. The fix is use the newly introduced CEPH_LOCK_FCNTL_INTR/CEPH_LOCK_FLOCK_INTR requests to interrupt blocked file lock request. These requests do not drop locks that have alread been acquired, they only interrupt blocked file lock request. Signed-off-by: Yan, Zheng <zyan@redhat.com>
2014-10-14 10:33:35 +08:00
/*
* wait for request completion callback
*/
typedef int (*ceph_mds_request_wait_callback_t) (struct ceph_mds_client *mdsc,
struct ceph_mds_request *req);
ceph: MDS client The MDS (metadata server) client is responsible for submitting requests to the MDS cluster and parsing the response. We decide which MDS to submit each request to based on cached information about the current partition of the directory hierarchy across the cluster. A stateful session is opened with each MDS before we submit requests to it, and a mutex is used to control the ordering of messages within each session. An MDS request may generate two responses. The first indicates the operation was a success and returns any result. A second reply is sent when the operation commits to disk. Note that locking on the MDS ensures that the results of updates are visible only to the updating client before the operation commits. Requests are linked to the containing directory so that an fsync will wait for them to commit. If an MDS fails and/or recovers, we resubmit requests as needed. We also reconnect existing capabilities to a recovering MDS to reestablish that shared session state. Old dentry leases are invalidated. Signed-off-by: Sage Weil <sage@newdream.net>
2009-10-07 02:31:09 +08:00
/*
* an in-flight mds request
*/
struct ceph_mds_request {
u64 r_tid; /* transaction id */
ceph: use rbtree for mds requests The rbtree is a more appropriate data structure than a radix_tree. It avoids extra memory usage and simplifies the code. It also fixes a bug where the debugfs 'mdsc' file wasn't including the most recent mds request. Signed-off-by: Sage Weil <sage@newdream.net>
2010-02-16 04:08:46 +08:00
struct rb_node r_node;
ceph: do caps accounting per mds_client Caps related accounting is now being done per mds client instead of just being global. This prepares ground work for a later revision of the caps preallocated reservation list. Signed-off-by: Yehuda Sadeh <yehuda@hq.newdream.net> Signed-off-by: Sage Weil <sage@newdream.net>
2010-06-18 07:16:12 +08:00
struct ceph_mds_client *r_mdsc;
ceph: MDS client The MDS (metadata server) client is responsible for submitting requests to the MDS cluster and parsing the response. We decide which MDS to submit each request to based on cached information about the current partition of the directory hierarchy across the cluster. A stateful session is opened with each MDS before we submit requests to it, and a mutex is used to control the ordering of messages within each session. An MDS request may generate two responses. The first indicates the operation was a success and returns any result. A second reply is sent when the operation commits to disk. Note that locking on the MDS ensures that the results of updates are visible only to the updating client before the operation commits. Requests are linked to the containing directory so that an fsync will wait for them to commit. If an MDS fails and/or recovers, we resubmit requests as needed. We also reconnect existing capabilities to a recovering MDS to reestablish that shared session state. Old dentry leases are invalidated. Signed-off-by: Sage Weil <sage@newdream.net>
2009-10-07 02:31:09 +08:00
int r_op; /* mds op code */
/* operation on what? */
struct inode *r_inode; /* arg1 */
struct dentry *r_dentry; /* arg1 */
struct dentry *r_old_dentry; /* arg2: rename from or link from */
ceph: explicitly reference rename old_dentry parent dir in request We carry a pin on the parent directory for the rename source and dest dentries. For the source it's r_locked_dir; we need to explicitly reference the old_dentry parent as well, since the dentry's d_parent may change between when the request was created and pinned and when it is freed. Reviewed-by: Yehuda Sadeh <yehuda@hq.newdream.net> Signed-off-by: Sage Weil <sage@newdream.net>
2011-07-27 02:31:14 +08:00
struct inode *r_old_dentry_dir; /* arg2: old dentry's parent dir */
ceph: MDS client The MDS (metadata server) client is responsible for submitting requests to the MDS cluster and parsing the response. We decide which MDS to submit each request to based on cached information about the current partition of the directory hierarchy across the cluster. A stateful session is opened with each MDS before we submit requests to it, and a mutex is used to control the ordering of messages within each session. An MDS request may generate two responses. The first indicates the operation was a success and returns any result. A second reply is sent when the operation commits to disk. Note that locking on the MDS ensures that the results of updates are visible only to the updating client before the operation commits. Requests are linked to the containing directory so that an fsync will wait for them to commit. If an MDS fails and/or recovers, we resubmit requests as needed. We also reconnect existing capabilities to a recovering MDS to reestablish that shared session state. Old dentry leases are invalidated. Signed-off-by: Sage Weil <sage@newdream.net>
2009-10-07 02:31:09 +08:00
char *r_path1, *r_path2;
struct ceph_vino r_ino1, r_ino2;
struct inode *r_locked_dir; /* dir (if any) i_mutex locked by vfs */
struct inode *r_target_inode; /* resulting inode */
ceph: fix race between aborted requests and fill_trace When we abort requests we need to prevent fill_trace et al from doing anything that relies on locks held by the VFS caller. This fixes a race between the reply handler and the abort code, ensuring that continue holding the dir mutex until the reply handler completes. Signed-off-by: Sage Weil <sage@newdream.net>
2010-05-14 03:01:13 +08:00
struct mutex r_fill_mutex;
ceph: MDS client The MDS (metadata server) client is responsible for submitting requests to the MDS cluster and parsing the response. We decide which MDS to submit each request to based on cached information about the current partition of the directory hierarchy across the cluster. A stateful session is opened with each MDS before we submit requests to it, and a mutex is used to control the ordering of messages within each session. An MDS request may generate two responses. The first indicates the operation was a success and returns any result. A second reply is sent when the operation commits to disk. Note that locking on the MDS ensures that the results of updates are visible only to the updating client before the operation commits. Requests are linked to the containing directory so that an fsync will wait for them to commit. If an MDS fails and/or recovers, we resubmit requests as needed. We also reconnect existing capabilities to a recovering MDS to reestablish that shared session state. Old dentry leases are invalidated. Signed-off-by: Sage Weil <sage@newdream.net>
2009-10-07 02:31:09 +08:00
union ceph_mds_request_args r_args;
int r_fmode; /* file mode, if expecting cap */
ceph: Convert struct ceph_mds_request to use kuid_t and kgid_t Hold the uid and gid for a pending ceph mds request using the types kuid_t and kgid_t. When a request message is finally created convert the kuid_t and kgid_t values into uids and gids in the initial user namespace. Cc: Sage Weil <sage@inktank.com> Signed-off-by: "Eric W. Biederman" <ebiederm@xmission.com>
2013-01-31 20:01:53 +08:00
kuid_t r_uid;
kgid_t r_gid;
ceph: include time stamp in every MDS request We recently modified the client/MDS protocol to include a timestamp in the client request. This allows ctime updates to follow the client's clock in most cases, which avoids subtle problems when clocks are out of sync and timestamps are updated sometimes by the MDS clock (for most requests) and sometimes by the client clock (for cap writeback). Signed-off-by: Sage Weil <sage@inktank.com>
2014-05-22 08:41:08 +08:00
struct timespec r_stamp;
ceph: MDS client The MDS (metadata server) client is responsible for submitting requests to the MDS cluster and parsing the response. We decide which MDS to submit each request to based on cached information about the current partition of the directory hierarchy across the cluster. A stateful session is opened with each MDS before we submit requests to it, and a mutex is used to control the ordering of messages within each session. An MDS request may generate two responses. The first indicates the operation was a success and returns any result. A second reply is sent when the operation commits to disk. Note that locking on the MDS ensures that the results of updates are visible only to the updating client before the operation commits. Requests are linked to the containing directory so that an fsync will wait for them to commit. If an MDS fails and/or recovers, we resubmit requests as needed. We also reconnect existing capabilities to a recovering MDS to reestablish that shared session state. Old dentry leases are invalidated. Signed-off-by: Sage Weil <sage@newdream.net>
2009-10-07 02:31:09 +08:00
/* for choosing which mds to send this request to */
int r_direct_mode;
u32 r_direct_hash; /* choose dir frag based on this dentry hash */
bool r_direct_is_hash; /* true if r_direct_hash is valid */
/* data payload is used for xattr ops */
ceph: use pagelist to present MDS request data Current code uses page array to present MDS request data. Pages in the array are allocated/freed by caller of ceph_mdsc_do_request(). If request is interrupted, the pages can be freed while they are still being used by the request message. The fix is use pagelist to present MDS request data. Pagelist is reference counted. Signed-off-by: Yan, Zheng <zyan@redhat.com> Reviewed-by: Sage Weil <sage@redhat.com>
2014-09-16 19:15:28 +08:00
struct ceph_pagelist *r_pagelist;
ceph: MDS client The MDS (metadata server) client is responsible for submitting requests to the MDS cluster and parsing the response. We decide which MDS to submit each request to based on cached information about the current partition of the directory hierarchy across the cluster. A stateful session is opened with each MDS before we submit requests to it, and a mutex is used to control the ordering of messages within each session. An MDS request may generate two responses. The first indicates the operation was a success and returns any result. A second reply is sent when the operation commits to disk. Note that locking on the MDS ensures that the results of updates are visible only to the updating client before the operation commits. Requests are linked to the containing directory so that an fsync will wait for them to commit. If an MDS fails and/or recovers, we resubmit requests as needed. We also reconnect existing capabilities to a recovering MDS to reestablish that shared session state. Old dentry leases are invalidated. Signed-off-by: Sage Weil <sage@newdream.net>
2009-10-07 02:31:09 +08:00
/* what caps shall we drop? */
int r_inode_drop, r_inode_unless;
int r_dentry_drop, r_dentry_unless;
int r_old_dentry_drop, r_old_dentry_unless;
struct inode *r_old_inode;
int r_old_inode_drop, r_old_inode_unless;
struct ceph_msg *r_request; /* original request */
ceph: do not include cap/dentry releases in replayed messages Strip the cap and dentry releases from replayed messages. They can cause the shared state to get out of sync because they were generated (with the request message) earlier, and no longer reflect the current client state. Signed-off-by: Sage Weil <sage@newdream.net>
2010-07-16 05:58:39 +08:00
int r_request_release_offset;
ceph: MDS client The MDS (metadata server) client is responsible for submitting requests to the MDS cluster and parsing the response. We decide which MDS to submit each request to based on cached information about the current partition of the directory hierarchy across the cluster. A stateful session is opened with each MDS before we submit requests to it, and a mutex is used to control the ordering of messages within each session. An MDS request may generate two responses. The first indicates the operation was a success and returns any result. A second reply is sent when the operation commits to disk. Note that locking on the MDS ensures that the results of updates are visible only to the updating client before the operation commits. Requests are linked to the containing directory so that an fsync will wait for them to commit. If an MDS fails and/or recovers, we resubmit requests as needed. We also reconnect existing capabilities to a recovering MDS to reestablish that shared session state. Old dentry leases are invalidated. Signed-off-by: Sage Weil <sage@newdream.net>
2009-10-07 02:31:09 +08:00
struct ceph_msg *r_reply;
struct ceph_mds_reply_info_parsed r_reply_info;
ceph: use getattr request to fetch inline data Add a new parameter 'locked_page' to ceph_do_getattr(). If inline data in getattr reply will be copied to the page. Signed-off-by: Yan, Zheng <zyan@redhat.com>
2014-11-14 21:56:29 +08:00
struct page *r_locked_page;
ceph: MDS client The MDS (metadata server) client is responsible for submitting requests to the MDS cluster and parsing the response. We decide which MDS to submit each request to based on cached information about the current partition of the directory hierarchy across the cluster. A stateful session is opened with each MDS before we submit requests to it, and a mutex is used to control the ordering of messages within each session. An MDS request may generate two responses. The first indicates the operation was a success and returns any result. A second reply is sent when the operation commits to disk. Note that locking on the MDS ensures that the results of updates are visible only to the updating client before the operation commits. Requests are linked to the containing directory so that an fsync will wait for them to commit. If an MDS fails and/or recovers, we resubmit requests as needed. We also reconnect existing capabilities to a recovering MDS to reestablish that shared session state. Old dentry leases are invalidated. Signed-off-by: Sage Weil <sage@newdream.net>
2009-10-07 02:31:09 +08:00
int r_err;
ceph: properly handle aborted mds requests Previously, if the MDS request was interrupted, we would unregister the request and ignore any reply. This could cause the caps or other cache state to become out of sync. (For instance, aborting dbench and doing rm -r on clients would complain about a non-empty directory because the client didn't realize it's aborted file create request completed.) Even we don't unregister, we still can't process the reply normally because we are no longer holding the caller's locks (like the dir i_mutex). So, mark aborted operations with r_aborted, and in the reply handler, be sure to process all the caps. Do not process the namespace changes, though, since we no longer will hold the dir i_mutex. The dentry lease state can also be ignored as it's more forgiving. Signed-off-by: Sage Weil <sage@newdream.net>
2010-01-26 03:33:08 +08:00
bool r_aborted;
ceph: MDS client The MDS (metadata server) client is responsible for submitting requests to the MDS cluster and parsing the response. We decide which MDS to submit each request to based on cached information about the current partition of the directory hierarchy across the cluster. A stateful session is opened with each MDS before we submit requests to it, and a mutex is used to control the ordering of messages within each session. An MDS request may generate two responses. The first indicates the operation was a success and returns any result. A second reply is sent when the operation commits to disk. Note that locking on the MDS ensures that the results of updates are visible only to the updating client before the operation commits. Requests are linked to the containing directory so that an fsync will wait for them to commit. If an MDS fails and/or recovers, we resubmit requests as needed. We also reconnect existing capabilities to a recovering MDS to reestablish that shared session state. Old dentry leases are invalidated. Signed-off-by: Sage Weil <sage@newdream.net>
2009-10-07 02:31:09 +08:00
unsigned long r_timeout; /* optional. jiffies */
unsigned long r_started; /* start time to measure timeout against */
unsigned long r_request_started; /* start time for mds request only,
used to measure lease durations */
/* link unsafe requests to parent directory, for fsync */
struct inode *r_unsafe_dir;
struct list_head r_unsafe_dir_item;
struct ceph_mds_session *r_session;
int r_attempts; /* resend attempts */
int r_num_fwd; /* number of forward attempts */
int r_resend_mds; /* mds to resend to next, if any*/
ceph: handle ESTALE properly; on receipt send to authority if it wasn't Signed-off-by: Greg Farnum <gregf@hq.newdream.net> Signed-off-by: Sage Weil <sage@newdream.net>
2010-06-23 06:58:01 +08:00
u32 r_sent_on_mseq; /* cap mseq request was sent at*/
ceph: MDS client The MDS (metadata server) client is responsible for submitting requests to the MDS cluster and parsing the response. We decide which MDS to submit each request to based on cached information about the current partition of the directory hierarchy across the cluster. A stateful session is opened with each MDS before we submit requests to it, and a mutex is used to control the ordering of messages within each session. An MDS request may generate two responses. The first indicates the operation was a success and returns any result. A second reply is sent when the operation commits to disk. Note that locking on the MDS ensures that the results of updates are visible only to the updating client before the operation commits. Requests are linked to the containing directory so that an fsync will wait for them to commit. If an MDS fails and/or recovers, we resubmit requests as needed. We also reconnect existing capabilities to a recovering MDS to reestablish that shared session state. Old dentry leases are invalidated. Signed-off-by: Sage Weil <sage@newdream.net>
2009-10-07 02:31:09 +08:00
ceph: use kref for struct ceph_mds_request Signed-off-by: Sage Weil <sage@newdream.net>
2009-12-08 04:31:09 +08:00
struct kref r_kref;
ceph: MDS client The MDS (metadata server) client is responsible for submitting requests to the MDS cluster and parsing the response. We decide which MDS to submit each request to based on cached information about the current partition of the directory hierarchy across the cluster. A stateful session is opened with each MDS before we submit requests to it, and a mutex is used to control the ordering of messages within each session. An MDS request may generate two responses. The first indicates the operation was a success and returns any result. A second reply is sent when the operation commits to disk. Note that locking on the MDS ensures that the results of updates are visible only to the updating client before the operation commits. Requests are linked to the containing directory so that an fsync will wait for them to commit. If an MDS fails and/or recovers, we resubmit requests as needed. We also reconnect existing capabilities to a recovering MDS to reestablish that shared session state. Old dentry leases are invalidated. Signed-off-by: Sage Weil <sage@newdream.net>
2009-10-07 02:31:09 +08:00
struct list_head r_wait;
struct completion r_completion;
struct completion r_safe_completion;
ceph_mds_request_callback_t r_callback;
ceph: fix file lock interruption When a lock operation is interrupted, current code sends a unlock request to MDS to undo the lock operation. This method does not work as expected because the unlock request can drop locks that have already been acquired. The fix is use the newly introduced CEPH_LOCK_FCNTL_INTR/CEPH_LOCK_FLOCK_INTR requests to interrupt blocked file lock request. These requests do not drop locks that have alread been acquired, they only interrupt blocked file lock request. Signed-off-by: Yan, Zheng <zyan@redhat.com>
2014-10-14 10:33:35 +08:00
ceph_mds_request_wait_callback_t r_wait_for_completion;
ceph: MDS client The MDS (metadata server) client is responsible for submitting requests to the MDS cluster and parsing the response. We decide which MDS to submit each request to based on cached information about the current partition of the directory hierarchy across the cluster. A stateful session is opened with each MDS before we submit requests to it, and a mutex is used to control the ordering of messages within each session. An MDS request may generate two responses. The first indicates the operation was a success and returns any result. A second reply is sent when the operation commits to disk. Note that locking on the MDS ensures that the results of updates are visible only to the updating client before the operation commits. Requests are linked to the containing directory so that an fsync will wait for them to commit. If an MDS fails and/or recovers, we resubmit requests as needed. We also reconnect existing capabilities to a recovering MDS to reestablish that shared session state. Old dentry leases are invalidated. Signed-off-by: Sage Weil <sage@newdream.net>
2009-10-07 02:31:09 +08:00
struct list_head r_unsafe_item; /* per-session unsafe list item */
ceph: clean up mds reply, error handling We would occasionally BUG out in the reply handler because r_reply was nonzero, due to a race with ceph_mdsc_do_request temporarily setting r_reply to an ERR_PTR value. This is unnecessary, messy, and also wrong in the EIO case. Clean up by consistently using r_err for errors and r_reply for messages. Also fix the abort logic to trigger consistently for all errors that return to the caller early (e.g., EIO from timeout case). If an abort races with a reply, use the result from the reply. Also fix locking for r_err, r_reply update in the reply handler. Signed-off-by: Sage Weil <sage@newdream.net>
2010-05-14 02:19:06 +08:00
bool r_got_unsafe, r_got_safe, r_got_result;
ceph: MDS client The MDS (metadata server) client is responsible for submitting requests to the MDS cluster and parsing the response. We decide which MDS to submit each request to based on cached information about the current partition of the directory hierarchy across the cluster. A stateful session is opened with each MDS before we submit requests to it, and a mutex is used to control the ordering of messages within each session. An MDS request may generate two responses. The first indicates the operation was a success and returns any result. A second reply is sent when the operation commits to disk. Note that locking on the MDS ensures that the results of updates are visible only to the updating client before the operation commits. Requests are linked to the containing directory so that an fsync will wait for them to commit. If an MDS fails and/or recovers, we resubmit requests as needed. We also reconnect existing capabilities to a recovering MDS to reestablish that shared session state. Old dentry leases are invalidated. Signed-off-by: Sage Weil <sage@newdream.net>
2009-10-07 02:31:09 +08:00
bool r_did_prepopulate;
u32 r_readdir_offset;
struct ceph_cap_reservation r_caps_reservation;
int r_num_caps;
};
/*
* mds client state
*/
struct ceph_mds_client {
ceph: factor out libceph from Ceph file system This factors out protocol and low-level storage parts of ceph into a separate libceph module living in net/ceph and include/linux/ceph. This is mostly a matter of moving files around. However, a few key pieces of the interface change as well: - ceph_client becomes ceph_fs_client and ceph_client, where the latter captures the mon and osd clients, and the fs_client gets the mds client and file system specific pieces. - Mount option parsing and debugfs setup is correspondingly broken into two pieces. - The mon client gets a generic handler callback for otherwise unknown messages (mds map, in this case). - The basic supported/required feature bits can be expanded (and are by ceph_fs_client). No functional change, aside from some subtle error handling cases that got cleaned up in the refactoring process. Signed-off-by: Sage Weil <sage@newdream.net>
2010-04-07 06:14:15 +08:00
struct ceph_fs_client *fsc;
ceph: MDS client The MDS (metadata server) client is responsible for submitting requests to the MDS cluster and parsing the response. We decide which MDS to submit each request to based on cached information about the current partition of the directory hierarchy across the cluster. A stateful session is opened with each MDS before we submit requests to it, and a mutex is used to control the ordering of messages within each session. An MDS request may generate two responses. The first indicates the operation was a success and returns any result. A second reply is sent when the operation commits to disk. Note that locking on the MDS ensures that the results of updates are visible only to the updating client before the operation commits. Requests are linked to the containing directory so that an fsync will wait for them to commit. If an MDS fails and/or recovers, we resubmit requests as needed. We also reconnect existing capabilities to a recovering MDS to reestablish that shared session state. Old dentry leases are invalidated. Signed-off-by: Sage Weil <sage@newdream.net>
2009-10-07 02:31:09 +08:00
struct mutex mutex; /* all nested structures */
struct ceph_mdsmap *mdsmap;
ceph: fix multiple mds session shutdown The use of a completion when waiting for session shutdown during umount is inappropriate, given the complexity of the condition. For multiple MDS's, this resulted in the umount thread spinning, often preventing the session close message from being processed in some cases. Switch to a waitqueue and defined a condition helper. This cleans things up nicely. Signed-off-by: Sage Weil <sage@newdream.net>
2010-08-12 05:51:23 +08:00
struct completion safe_umount_waiters;
wait_queue_head_t session_close_wq;
ceph: MDS client The MDS (metadata server) client is responsible for submitting requests to the MDS cluster and parsing the response. We decide which MDS to submit each request to based on cached information about the current partition of the directory hierarchy across the cluster. A stateful session is opened with each MDS before we submit requests to it, and a mutex is used to control the ordering of messages within each session. An MDS request may generate two responses. The first indicates the operation was a success and returns any result. A second reply is sent when the operation commits to disk. Note that locking on the MDS ensures that the results of updates are visible only to the updating client before the operation commits. Requests are linked to the containing directory so that an fsync will wait for them to commit. If an MDS fails and/or recovers, we resubmit requests as needed. We also reconnect existing capabilities to a recovering MDS to reestablish that shared session state. Old dentry leases are invalidated. Signed-off-by: Sage Weil <sage@newdream.net>
2009-10-07 02:31:09 +08:00
struct list_head waiting_for_map;
struct ceph_mds_session **sessions; /* NULL for mds if no session */
ceph: avoid block operation when !TASK_RUNNING (ceph_mdsc_close_sessions) use an atomic variable to track number of sessions, this can avoid block operation inside wait loops. Signed-off-by: Yan, Zheng <zyan@redhat.com>
2015-01-09 17:00:42 +08:00
atomic_t num_sessions;
ceph: MDS client The MDS (metadata server) client is responsible for submitting requests to the MDS cluster and parsing the response. We decide which MDS to submit each request to based on cached information about the current partition of the directory hierarchy across the cluster. A stateful session is opened with each MDS before we submit requests to it, and a mutex is used to control the ordering of messages within each session. An MDS request may generate two responses. The first indicates the operation was a success and returns any result. A second reply is sent when the operation commits to disk. Note that locking on the MDS ensures that the results of updates are visible only to the updating client before the operation commits. Requests are linked to the containing directory so that an fsync will wait for them to commit. If an MDS fails and/or recovers, we resubmit requests as needed. We also reconnect existing capabilities to a recovering MDS to reestablish that shared session state. Old dentry leases are invalidated. Signed-off-by: Sage Weil <sage@newdream.net>
2009-10-07 02:31:09 +08:00
int max_sessions; /* len of s_mds_sessions */
int stopping; /* true if shutting down */
/*
* snap_rwsem will cover cap linkage into snaprealms, and
* realm snap contexts. (later, we can do per-realm snap
* contexts locks..) the empty list contains realms with no
* references (implying they contain no inodes with caps) that
* should be destroyed.
*/
struct rw_semaphore snap_rwsem;
ceph: use rbtree for snap_realms Switch from radix tree to rbtree for snap realms. This is much more appropriate given that realm keys are few and far between. Signed-off-by: Sage Weil <sage@newdream.net>
2010-02-16 06:37:55 +08:00
struct rb_root snap_realms;
ceph: MDS client The MDS (metadata server) client is responsible for submitting requests to the MDS cluster and parsing the response. We decide which MDS to submit each request to based on cached information about the current partition of the directory hierarchy across the cluster. A stateful session is opened with each MDS before we submit requests to it, and a mutex is used to control the ordering of messages within each session. An MDS request may generate two responses. The first indicates the operation was a success and returns any result. A second reply is sent when the operation commits to disk. Note that locking on the MDS ensures that the results of updates are visible only to the updating client before the operation commits. Requests are linked to the containing directory so that an fsync will wait for them to commit. If an MDS fails and/or recovers, we resubmit requests as needed. We also reconnect existing capabilities to a recovering MDS to reestablish that shared session state. Old dentry leases are invalidated. Signed-off-by: Sage Weil <sage@newdream.net>
2009-10-07 02:31:09 +08:00
struct list_head snap_empty;
spinlock_t snap_empty_lock; /* protect snap_empty */
u64 last_tid; /* most recent mds request */
ceph: use rbtree for mds requests The rbtree is a more appropriate data structure than a radix_tree. It avoids extra memory usage and simplifies the code. It also fixes a bug where the debugfs 'mdsc' file wasn't including the most recent mds request. Signed-off-by: Sage Weil <sage@newdream.net>
2010-02-16 04:08:46 +08:00
struct rb_root request_tree; /* pending mds requests */
ceph: MDS client The MDS (metadata server) client is responsible for submitting requests to the MDS cluster and parsing the response. We decide which MDS to submit each request to based on cached information about the current partition of the directory hierarchy across the cluster. A stateful session is opened with each MDS before we submit requests to it, and a mutex is used to control the ordering of messages within each session. An MDS request may generate two responses. The first indicates the operation was a success and returns any result. A second reply is sent when the operation commits to disk. Note that locking on the MDS ensures that the results of updates are visible only to the updating client before the operation commits. Requests are linked to the containing directory so that an fsync will wait for them to commit. If an MDS fails and/or recovers, we resubmit requests as needed. We also reconnect existing capabilities to a recovering MDS to reestablish that shared session state. Old dentry leases are invalidated. Signed-off-by: Sage Weil <sage@newdream.net>
2009-10-07 02:31:09 +08:00
struct delayed_work delayed_work; /* delayed work */
unsigned long last_renew_caps; /* last time we renewed our caps */
struct list_head cap_delay_list; /* caps with delayed release */
spinlock_t cap_delay_lock; /* protects cap_delay_list */
struct list_head snap_flush_list; /* cap_snaps ready to flush */
spinlock_t snap_flush_lock;
u64 cap_flush_seq;
struct list_head cap_dirty; /* inodes with dirty caps */
ceph: fix cap flush race reentrancy In e9964c10 we change cap flushing to do a delicate dance because some inodes on the cap_dirty list could be in a migrating state (got EXPORT but not IMPORT) in which we couldn't actually flush and move from dirty->flushing, breaking the while (!empty) { process first } loop structure. It worked for a single sync thread, but was not reentrant and triggered infinite loops when multiple syncers came along. Instead, move inodes with dirty to a separate cap_dirty_migrating list when in the limbo export-but-no-import state, allowing us to go back to the simple loop structure (which was reentrant). This is cleaner and more robust. Audited the cap_dirty users and this looks fine: list_empty(&ci->i_dirty_item) is still a reliable indicator of whether we have dirty caps (which list we're on is irrelevant) and list_del_init() calls still do the right thing. Signed-off-by: Sage Weil <sage@newdream.net>
2011-05-25 02:46:31 +08:00
struct list_head cap_dirty_migrating; /* ...that are migration... */
ceph: MDS client The MDS (metadata server) client is responsible for submitting requests to the MDS cluster and parsing the response. We decide which MDS to submit each request to based on cached information about the current partition of the directory hierarchy across the cluster. A stateful session is opened with each MDS before we submit requests to it, and a mutex is used to control the ordering of messages within each session. An MDS request may generate two responses. The first indicates the operation was a success and returns any result. A second reply is sent when the operation commits to disk. Note that locking on the MDS ensures that the results of updates are visible only to the updating client before the operation commits. Requests are linked to the containing directory so that an fsync will wait for them to commit. If an MDS fails and/or recovers, we resubmit requests as needed. We also reconnect existing capabilities to a recovering MDS to reestablish that shared session state. Old dentry leases are invalidated. Signed-off-by: Sage Weil <sage@newdream.net>
2009-10-07 02:31:09 +08:00
int num_cap_flushing; /* # caps we are flushing */
spinlock_t cap_dirty_lock; /* protects above items */
wait_queue_head_t cap_flushing_wq;
ceph: do caps accounting per mds_client Caps related accounting is now being done per mds client instead of just being global. This prepares ground work for a later revision of the caps preallocated reservation list. Signed-off-by: Yehuda Sadeh <yehuda@hq.newdream.net> Signed-off-by: Sage Weil <sage@newdream.net>
2010-06-18 07:16:12 +08:00
/*
* Cap reservations
*
* Maintain a global pool of preallocated struct ceph_caps, referenced
* by struct ceph_caps_reservations. This ensures that we preallocate
* memory needed to successfully process an MDS response. (If an MDS
* sends us cap information and we fail to process it, we will have
* problems due to the client and MDS being out of sync.)
*
* Reservations are 'owned' by a ceph_cap_reservation context.
*/
spinlock_t caps_list_lock;
struct list_head caps_list; /* unused (reserved or
unreserved) */
int caps_total_count; /* total caps allocated */
int caps_use_count; /* in use */
int caps_reserve_count; /* unused, reserved */
int caps_avail_count; /* unused, unreserved */
int caps_min_count; /* keep at least this many
(unreserved) */
ceph: MDS client The MDS (metadata server) client is responsible for submitting requests to the MDS cluster and parsing the response. We decide which MDS to submit each request to based on cached information about the current partition of the directory hierarchy across the cluster. A stateful session is opened with each MDS before we submit requests to it, and a mutex is used to control the ordering of messages within each session. An MDS request may generate two responses. The first indicates the operation was a success and returns any result. A second reply is sent when the operation commits to disk. Note that locking on the MDS ensures that the results of updates are visible only to the updating client before the operation commits. Requests are linked to the containing directory so that an fsync will wait for them to commit. If an MDS fails and/or recovers, we resubmit requests as needed. We also reconnect existing capabilities to a recovering MDS to reestablish that shared session state. Old dentry leases are invalidated. Signed-off-by: Sage Weil <sage@newdream.net>
2009-10-07 02:31:09 +08:00
spinlock_t dentry_lru_lock;
struct list_head dentry_lru;
int num_dentry;
};
extern const char *ceph_mds_op_name(int op);
extern struct ceph_mds_session *
__ceph_lookup_mds_session(struct ceph_mds_client *, int mds);
static inline struct ceph_mds_session *
ceph_get_mds_session(struct ceph_mds_session *s)
{
atomic_inc(&s->s_ref);
return s;
}
ceph: export ceph_session_state_name function ...so that it can be used from the ceph debugfs code when dumping session info. Signed-off-by: John Spray <john.spray@redhat.com>
2014-09-19 20:51:08 +08:00
extern const char *ceph_session_state_name(int s);
ceph: MDS client The MDS (metadata server) client is responsible for submitting requests to the MDS cluster and parsing the response. We decide which MDS to submit each request to based on cached information about the current partition of the directory hierarchy across the cluster. A stateful session is opened with each MDS before we submit requests to it, and a mutex is used to control the ordering of messages within each session. An MDS request may generate two responses. The first indicates the operation was a success and returns any result. A second reply is sent when the operation commits to disk. Note that locking on the MDS ensures that the results of updates are visible only to the updating client before the operation commits. Requests are linked to the containing directory so that an fsync will wait for them to commit. If an MDS fails and/or recovers, we resubmit requests as needed. We also reconnect existing capabilities to a recovering MDS to reestablish that shared session state. Old dentry leases are invalidated. Signed-off-by: Sage Weil <sage@newdream.net>
2009-10-07 02:31:09 +08:00
extern void ceph_put_mds_session(struct ceph_mds_session *s);
extern int ceph_send_msg_mds(struct ceph_mds_client *mdsc,
struct ceph_msg *msg, int mds);
ceph: factor out libceph from Ceph file system This factors out protocol and low-level storage parts of ceph into a separate libceph module living in net/ceph and include/linux/ceph. This is mostly a matter of moving files around. However, a few key pieces of the interface change as well: - ceph_client becomes ceph_fs_client and ceph_client, where the latter captures the mon and osd clients, and the fs_client gets the mds client and file system specific pieces. - Mount option parsing and debugfs setup is correspondingly broken into two pieces. - The mon client gets a generic handler callback for otherwise unknown messages (mds map, in this case). - The basic supported/required feature bits can be expanded (and are by ceph_fs_client). No functional change, aside from some subtle error handling cases that got cleaned up in the refactoring process. Signed-off-by: Sage Weil <sage@newdream.net>
2010-04-07 06:14:15 +08:00
extern int ceph_mdsc_init(struct ceph_fs_client *fsc);
ceph: MDS client The MDS (metadata server) client is responsible for submitting requests to the MDS cluster and parsing the response. We decide which MDS to submit each request to based on cached information about the current partition of the directory hierarchy across the cluster. A stateful session is opened with each MDS before we submit requests to it, and a mutex is used to control the ordering of messages within each session. An MDS request may generate two responses. The first indicates the operation was a success and returns any result. A second reply is sent when the operation commits to disk. Note that locking on the MDS ensures that the results of updates are visible only to the updating client before the operation commits. Requests are linked to the containing directory so that an fsync will wait for them to commit. If an MDS fails and/or recovers, we resubmit requests as needed. We also reconnect existing capabilities to a recovering MDS to reestablish that shared session state. Old dentry leases are invalidated. Signed-off-by: Sage Weil <sage@newdream.net>
2009-10-07 02:31:09 +08:00
extern void ceph_mdsc_close_sessions(struct ceph_mds_client *mdsc);
ceph: factor out libceph from Ceph file system This factors out protocol and low-level storage parts of ceph into a separate libceph module living in net/ceph and include/linux/ceph. This is mostly a matter of moving files around. However, a few key pieces of the interface change as well: - ceph_client becomes ceph_fs_client and ceph_client, where the latter captures the mon and osd clients, and the fs_client gets the mds client and file system specific pieces. - Mount option parsing and debugfs setup is correspondingly broken into two pieces. - The mon client gets a generic handler callback for otherwise unknown messages (mds map, in this case). - The basic supported/required feature bits can be expanded (and are by ceph_fs_client). No functional change, aside from some subtle error handling cases that got cleaned up in the refactoring process. Signed-off-by: Sage Weil <sage@newdream.net>
2010-04-07 06:14:15 +08:00
extern void ceph_mdsc_destroy(struct ceph_fs_client *fsc);
ceph: MDS client The MDS (metadata server) client is responsible for submitting requests to the MDS cluster and parsing the response. We decide which MDS to submit each request to based on cached information about the current partition of the directory hierarchy across the cluster. A stateful session is opened with each MDS before we submit requests to it, and a mutex is used to control the ordering of messages within each session. An MDS request may generate two responses. The first indicates the operation was a success and returns any result. A second reply is sent when the operation commits to disk. Note that locking on the MDS ensures that the results of updates are visible only to the updating client before the operation commits. Requests are linked to the containing directory so that an fsync will wait for them to commit. If an MDS fails and/or recovers, we resubmit requests as needed. We also reconnect existing capabilities to a recovering MDS to reestablish that shared session state. Old dentry leases are invalidated. Signed-off-by: Sage Weil <sage@newdream.net>
2009-10-07 02:31:09 +08:00
extern void ceph_mdsc_sync(struct ceph_mds_client *mdsc);
extern void ceph_mdsc_lease_release(struct ceph_mds_client *mdsc,
struct inode *inode,
ceph: ignore lease mask The lease mask is no longer used (and it changed a while back). Instead, use a non-zero duration to indicate that there is a lease being issued. Reviewed-by: Yehuda Sadeh <yehuda@hq.newdream.net> Signed-off-by: Sage Weil <sage@newdream.net>
2011-07-27 02:28:25 +08:00
struct dentry *dn);
ceph: MDS client The MDS (metadata server) client is responsible for submitting requests to the MDS cluster and parsing the response. We decide which MDS to submit each request to based on cached information about the current partition of the directory hierarchy across the cluster. A stateful session is opened with each MDS before we submit requests to it, and a mutex is used to control the ordering of messages within each session. An MDS request may generate two responses. The first indicates the operation was a success and returns any result. A second reply is sent when the operation commits to disk. Note that locking on the MDS ensures that the results of updates are visible only to the updating client before the operation commits. Requests are linked to the containing directory so that an fsync will wait for them to commit. If an MDS fails and/or recovers, we resubmit requests as needed. We also reconnect existing capabilities to a recovering MDS to reestablish that shared session state. Old dentry leases are invalidated. Signed-off-by: Sage Weil <sage@newdream.net>
2009-10-07 02:31:09 +08:00
ceph: use common helper for aborted dir request invalidation We invalidate I_COMPLETE and dentry leases in two places: on aborted mds request and on request replay. Use common helper to avoid duplicate code. Signed-off-by: Sage Weil <sage@newdream.net>
2010-05-15 01:02:57 +08:00
extern void ceph_invalidate_dir_request(struct ceph_mds_request *req);
ceph: preallocate buffer for readdir reply Preallocate buffer for readdir reply. Limit number of entries in readdir reply according to the buffer size. Signed-off-by: Yan, Zheng <zheng.z.yan@intel.com>
2014-03-29 13:41:15 +08:00
extern int ceph_alloc_readdir_reply_buffer(struct ceph_mds_request *req,
struct inode *dir);
ceph: MDS client The MDS (metadata server) client is responsible for submitting requests to the MDS cluster and parsing the response. We decide which MDS to submit each request to based on cached information about the current partition of the directory hierarchy across the cluster. A stateful session is opened with each MDS before we submit requests to it, and a mutex is used to control the ordering of messages within each session. An MDS request may generate two responses. The first indicates the operation was a success and returns any result. A second reply is sent when the operation commits to disk. Note that locking on the MDS ensures that the results of updates are visible only to the updating client before the operation commits. Requests are linked to the containing directory so that an fsync will wait for them to commit. If an MDS fails and/or recovers, we resubmit requests as needed. We also reconnect existing capabilities to a recovering MDS to reestablish that shared session state. Old dentry leases are invalidated. Signed-off-by: Sage Weil <sage@newdream.net>
2009-10-07 02:31:09 +08:00
extern struct ceph_mds_request *
ceph_mdsc_create_request(struct ceph_mds_client *mdsc, int op, int mode);
extern void ceph_mdsc_submit_request(struct ceph_mds_client *mdsc,
struct ceph_mds_request *req);
extern int ceph_mdsc_do_request(struct ceph_mds_client *mdsc,
struct inode *dir,
struct ceph_mds_request *req);
static inline void ceph_mdsc_get_request(struct ceph_mds_request *req)
{
ceph: use kref for struct ceph_mds_request Signed-off-by: Sage Weil <sage@newdream.net>
2009-12-08 04:31:09 +08:00
kref_get(&req->r_kref);
}
extern void ceph_mdsc_release_request(struct kref *kref);
static inline void ceph_mdsc_put_request(struct ceph_mds_request *req)
{
kref_put(&req->r_kref, ceph_mdsc_release_request);
ceph: MDS client The MDS (metadata server) client is responsible for submitting requests to the MDS cluster and parsing the response. We decide which MDS to submit each request to based on cached information about the current partition of the directory hierarchy across the cluster. A stateful session is opened with each MDS before we submit requests to it, and a mutex is used to control the ordering of messages within each session. An MDS request may generate two responses. The first indicates the operation was a success and returns any result. A second reply is sent when the operation commits to disk. Note that locking on the MDS ensures that the results of updates are visible only to the updating client before the operation commits. Requests are linked to the containing directory so that an fsync will wait for them to commit. If an MDS fails and/or recovers, we resubmit requests as needed. We also reconnect existing capabilities to a recovering MDS to reestablish that shared session state. Old dentry leases are invalidated. Signed-off-by: Sage Weil <sage@newdream.net>
2009-10-07 02:31:09 +08:00
}
ceph: try to send partial cap release on cap message on missing inode If we have enough memory to allocate a new cap release message, do so, so that we can send a partial release message immediately. This keeps us from making the MDS wait when the cap release it needs is in a partially full release message. If we fail because of ENOMEM, oh well, they'll just have to wait a bit longer. Signed-off-by: Sage Weil <sage@newdream.net>
2010-06-10 07:52:04 +08:00
extern int ceph_add_cap_releases(struct ceph_mds_client *mdsc,
ceph: drop unused argument Signed-off-by: Sage Weil <sage@newdream.net>
2010-06-11 03:55:52 +08:00
struct ceph_mds_session *session);
ceph: release cap on import if we don't have the inode If we get an IMPORT that give us a cap, but we don't have the inode, queue a release (and try to send it immediately) so that the MDS doesn't get stuck waiting for us. Signed-off-by: Sage Weil <sage@newdream.net>
2010-06-10 07:47:10 +08:00
extern void ceph_send_cap_releases(struct ceph_mds_client *mdsc,
struct ceph_mds_session *session);
ceph: MDS client The MDS (metadata server) client is responsible for submitting requests to the MDS cluster and parsing the response. We decide which MDS to submit each request to based on cached information about the current partition of the directory hierarchy across the cluster. A stateful session is opened with each MDS before we submit requests to it, and a mutex is used to control the ordering of messages within each session. An MDS request may generate two responses. The first indicates the operation was a success and returns any result. A second reply is sent when the operation commits to disk. Note that locking on the MDS ensures that the results of updates are visible only to the updating client before the operation commits. Requests are linked to the containing directory so that an fsync will wait for them to commit. If an MDS fails and/or recovers, we resubmit requests as needed. We also reconnect existing capabilities to a recovering MDS to reestablish that shared session state. Old dentry leases are invalidated. Signed-off-by: Sage Weil <sage@newdream.net>
2009-10-07 02:31:09 +08:00
extern void ceph_mdsc_pre_umount(struct ceph_mds_client *mdsc);
extern char *ceph_mdsc_build_path(struct dentry *dentry, int *plen, u64 *base,
int stop_on_nosnap);
extern void __ceph_mdsc_drop_dentry_lease(struct dentry *dentry);
extern void ceph_mdsc_lease_send_msg(struct ceph_mds_session *session,
struct inode *inode,
struct dentry *dentry, char action,
u32 seq);
extern void ceph_mdsc_handle_map(struct ceph_mds_client *mdsc,
struct ceph_msg *msg);
ceph: add open export target session helper Signed-off-by: Yan, Zheng <zheng.z.yan@intel.com>
2013-11-24 14:33:01 +08:00
extern struct ceph_mds_session *
ceph_mdsc_open_export_target_session(struct ceph_mds_client *mdsc, int target);
ceph: connect to export targets on cap export When we get a cap EXPORT message, make sure we are connected to all export targets to ensure we can handle the matching IMPORT. Signed-off-by: Sage Weil <sage@newdream.net>
2010-06-22 04:45:04 +08:00
extern void ceph_mdsc_open_export_target_sessions(struct ceph_mds_client *mdsc,
struct ceph_mds_session *session);
ceph: MDS client The MDS (metadata server) client is responsible for submitting requests to the MDS cluster and parsing the response. We decide which MDS to submit each request to based on cached information about the current partition of the directory hierarchy across the cluster. A stateful session is opened with each MDS before we submit requests to it, and a mutex is used to control the ordering of messages within each session. An MDS request may generate two responses. The first indicates the operation was a success and returns any result. A second reply is sent when the operation commits to disk. Note that locking on the MDS ensures that the results of updates are visible only to the updating client before the operation commits. Requests are linked to the containing directory so that an fsync will wait for them to commit. If an MDS fails and/or recovers, we resubmit requests as needed. We also reconnect existing capabilities to a recovering MDS to reestablish that shared session state. Old dentry leases are invalidated. Signed-off-by: Sage Weil <sage@newdream.net>
2009-10-07 02:31:09 +08:00
#endif