1. OSD::handle_op: validates permissions and crush mapping.
discard the request if they are not connected and the client cannot get the reply ( See OSD::op_is_discardable )
See OSDService::handle_misdirected_op
- See OSD::op_has_sufficient_caps
+ See PG::op_has_sufficient_caps
See OSD::require_same_or_newer_map
2. OSD::enqueue_op
These are the same in cases where there was no triggering message. We discard
a peering message if the message's query_epoch if the PG in question has entered
-a new epoch (See PG::old_peering_event, PG::queue_peering_event). Notifies,
+a new epoch (See PG::old_peering_evt, PG::queue_peering_event). Notifies,
infos, notifies, and logs are all handled as PG::RecoveryMachine events and
are wrapped by PG::queue_* by PG::CephPeeringEvts, which include the created
state machine event along with epoch_sent and query_epoch in order to
To remove a snapshot, a request is made to the *Monitor* cluster to
add the snapshot id to the list of purged snaps (or to remove it from
the set of pool snaps in the case of *pool snaps*). In either case,
-the *PG* adds the snap to its *snaptrimq* for trimming.
+the *PG* adds the snap to its *snap_trimq* for trimming.
A clone can be removed when all of its snaps have been removed. In
order to determine which clones might need to be removed upon snap
This trimming is performed asynchronously by the snap_trim_wq while the
pg is clean and not scrubbing.
- #. The next snap in PG::snaptrimq is selected for trimming
+ #. The next snap in PG::snap_trimq is selected for trimming
#. We determine the next object for trimming out of PG::snap_mapper.
For each object, we create a log entry and repop updating the
object info and the snap set (including adjusting the overlaps).
* prefix. The list is guaranteed to have at least one item when the
* function returns true.
*
- * @in
* @param prefix the prefix to match against
- * @param objects return list of objects
+ * @param objects [out] return list of objects
* @returns true if there are any objects in the store which match
* the prefix, false if there are no more
*/
*
* @param major where to store the major version number
* @param minor where to store the minor version number
- * @param extra where to store the extra version number
+ * @param patch where to store the extra version number
*/
const char *ceph_version(int *major, int *minor, int *patch);
* @param path the path to the working directory to change into.
* @returns 0 on success, negative error code otherwise.
*/
-int ceph_chdir(struct ceph_mount_info *cmount, const char *s);
+int ceph_chdir(struct ceph_mount_info *cmount, const char *path);
/** @} fsops */
* Change the ownership of a file from an open file descriptor.
*
* @param cmount the ceph mount handle to use for performing the chown.
- * @param path the path of the file/directory to change the ownership of.
+ * @param fd the fd of the open file/directory to change the ownership of.
* @param uid the user id to set on the file/directory.
* @param gid the group id to set on the file/directory.
* @returns 0 on success or negative error code on failure.
* @param offset the offset to set the stream to
* @param whence the flag to indicate what type of seeking to perform:
* SEEK_SET: the offset is set to the given offset in the file.
- * SEEK_CUR: the offset is set to the current location plus @ref offset bytes.
- * SEEK_END: the offset is set to the end of the file plus @ref offset bytes.
+ * SEEK_CUR: the offset is set to the current location plus @e offset bytes.
+ * SEEK_END: the offset is set to the end of the file plus @e offset bytes.
* @returns 0 on success or a negative error code on failure.
*/
int64_t ceph_lseek(struct ceph_mount_info *cmount, int fd, int64_t offset, int whence);
* @param cmount the ceph mount handle to use for performing the read.
* @param fd the file descriptor of the open file to read from.
* @param buf the buffer to read data into
- * @param the initial size of the buffer
+ * @param size the initial size of the buffer
* @param offset the offset in the file to read from. If this value is negative, the
* function reads from the current offset of the file descriptor.
* @returns the number of bytes read into buf, or a negative error code on failure.
* anywhere within the stripe unit.
* @param addr the address of the OSD holding that stripe
* @param naddr the capacity of the address passed in.
- * @returns the size of the addressed filled into the @ref addr parameter, or a negative
+ * @returns the size of the addressed filled into the @e addr parameter, or a negative
* error code on failure.
*/
int ceph_get_file_stripe_address(struct ceph_mount_info *cmount, int fd, int64_t offset,
* @returns the number of items stored in the output array, or -ERANGE if the
* array is not large enough.
*/
-int ceph_get_file_extent_osds(struct ceph_mount_info *cmount, int fh,
+int ceph_get_file_extent_osds(struct ceph_mount_info *cmount, int fd,
int64_t offset, int64_t *length, int *osds, int nosds);
/**
* Get the capabilities currently issued to the client.
*
* @param cmount the ceph mount handle to use.
- * @param the path to the file
+ * @param path the path to the file
* @returns the current capabilities issued to this client
* for the file
*/
/** @} */
/**
- * @defgroup librados_h_operation_flags
+ * @defgroup librados_h_operation_flags Operation Flags
* Flags for rados_read_op_opeprate(), rados_write_op_operate(),
* rados_aio_read_op_operate(), and rados_aio_write_op_operate().
* See librados.hpp for details.
* Share configuration state with another rados_t instance.
*
* @param cluster where to store the handle
- * @param cct_ the existing configuration to use
+ * @param cct the existing configuration to use
* @returns 0 on success, negative error code on failure
*/
CEPH_RADOS_API int rados_create_with_context(rados_t *cluster,
* @param io the pool the object is in
* @param o the object to watch
* @param cookie where to store the internal id assigned to this watch
- * @param watchcb2 what to do when a notify is received on this object
+ * @param watchcb what to do when a notify is received on this object
* @param watcherrcb what to do when the watch session encounters an error
* @param arg opaque value to pass to the callback
* @returns 0 on success, negative error code on failure
/**
* Truncate an object
* @param write_op operation to add this action to
- * @offset Offset to truncate to
+ * @param offset Offset to truncate to
*/
CEPH_RADOS_API void rados_write_op_truncate(rados_write_op_t write_op,
uint64_t offset);
/**
* Zero part of an object
* @param write_op operation to add this action to
- * @offset Offset to zero
- * @len length to zero
+ * @param offset Offset to zero
+ * @param len length to zero
*/
CEPH_RADOS_API void rados_write_op_zero(rados_write_op_t write_op,
uint64_t offset,
* @param read_op operation to add this action to
* @param ver object version number
*/
-CEPH_RADOS_API void rados_read_op_assert_version(rados_read_op_t write_op, uint64_t ver);
+CEPH_RADOS_API void rados_read_op_assert_version(rados_read_op_t read_op, uint64_t ver);
/**
* Ensure that the an xattr satisfies a comparison
*
* @param read_op operation to add this action to
* @param offset offset to read from
- * @param buffer where to put the data
* @param len length of buffer
- * @param prval where to store the return value of this action
+ * @param buffer where to put the data
* @param bytes_read where to store the number of bytes read by this action
+ * @param prval where to store the return value of this action
*/
CEPH_RADOS_API void rados_read_op_read(rados_read_op_t read_op,
uint64_t offset,
size_t len,
- char *buf,
+ char *buffer,
size_t *bytes_read,
int *prval);
* @param read_op operation to add this action to
* @param start_after list keys starting after start_after
* @param filter_prefix list only keys beginning with filter_prefix
- * @parem max_return list no more than max_return key/value pairs
+ * @param max_return list no more than max_return key/value pairs
* @param iter where to store the iterator
* @param prval where to store the return value from this action
*/
*
* @param read_op operation to add this action to
* @param start_after list keys starting after start_after
- * @parem max_return list no more than max_return keys
+ * @param max_return list no more than max_return keys
* @param iter where to store the iterator
* @param prval where to store the return value from this action
*/
/**
* Perform a read operation synchronously
* @param read_op operation to perform
- * @io the ioctx that the object is in
- * @oid the object id
- * @flags flags to apply to the entire operation (LIBRADOS_OPERATION_*)
+ * @param io the ioctx that the object is in
+ * @param oid the object id
+ * @param flags flags to apply to the entire operation (LIBRADOS_OPERATION_*)
*/
CEPH_RADOS_API int rados_read_op_operate(rados_read_op_t read_op,
rados_ioctx_t io,
/**
* Perform a read operation asynchronously
* @param read_op operation to perform
- * @io the ioctx that the object is in
+ * @param io the ioctx that the object is in
* @param completion what to do when operation has been attempted
- * @oid the object id
- * @flags flags to apply to the entire operation (LIBRADOS_OPERATION_*)
+ * @param oid the object id
+ * @param flags flags to apply to the entire operation (LIBRADOS_OPERATION_*)
*/
CEPH_RADOS_API int rados_aio_read_op_operate(rados_read_op_t read_op,
rados_ioctx_t io,
* @returns -EBUSY if the lock is already held by another (client, cookie) pair
* @returns -EEXIST if the lock is already held by the same (client, cookie) pair
*/
-CEPH_RADOS_API int rados_lock_exclusive(rados_ioctx_t io, const char * o,
+CEPH_RADOS_API int rados_lock_exclusive(rados_ioctx_t io, const char * oid,
const char * name, const char * cookie,
const char * desc,
struct timeval * duration,
* Get up to max_return keys and values beginning after start_after
*
* @param start_after [in] list no keys smaller than start_after
- * @parem max_return [in] list no more than max_return key/value pairs
+ * @param max_return [in] list no more than max_return key/value pairs
* @param out_vals [out] place returned values in out_vals on completion
* @param prval [out] place error code in prval upon completion
*/
*
* @param start_after [in] list keys starting after start_after
* @param filter_prefix [in] list only keys beginning with filter_prefix
- * @parem max_return [in] list no more than max_return key/value pairs
+ * @param max_return [in] list no more than max_return key/value pairs
* @param out_vals [out] place returned values in out_vals on completion
* @param prval [out] place error code in prval upon completion
*/
* Get up to max_return keys beginning after start_after
*
* @param start_after [in] list keys starting after start_after
- * @parem max_return [in] list no more than max_return keys
+ * @param max_return [in] list no more than max_return keys
* @param out_keys [out] place returned values in out_keys on completion
* @param prval [out] place error code in prval upon completion
*/
/**
* get key/value pairs for specified keys
*
- * @param to_get [in] keys to get
- * @param out_vals [out] place key/value pairs found here on completion
+ * @param keys [in] keys to get
+ * @param map [out] place key/value pairs found here on completion
* @param prval [out] place error code in prval upon completion
*/
void omap_get_vals_by_keys(const std::set<std::string> &keys,
/**
* query dirty state of an object
*
- * @param out_dirty [out] pointer to resulting bool
+ * @param isdirty [out] pointer to resulting bool
* @param prval [out] place error code in prval upon completion
*/
void is_dirty(bool *isdirty, int *prval);
* The return value of the completion will be 0 on success, negative
* error code on failure.
*
- * @param io the context to operate in
* @param oid the name of the object
- * @param completion what to do when the remove is safe and complete
+ * @param c what to do when the remove is safe and complete
* @returns 0 on success, -EROFS if the io context specifies a snap_seq
* other than SNAP_HEAD
*/
* Already existing objects will be opened with their own layout.
*
* @param striper the targetted striper
- * @param stiper_unit the stripe_unit value of the new object layout
+ * @param stripe_unit the stripe_unit value of the new object layout
* @returns 0 on success, negative error code on failure
*/
int rados_striper_set_object_layout_stripe_unit(rados_striper_t striper,
* @note the truncation can not happen if any I/O is ongoing (it
* will return EBUSY). Identically, no I/O will be able to start
* during truncation (same EBUSY return code)
- * @param striper the striper in which the truncation will occur
+ * @param io the rados context to use
* @param soid the name of the striped object
* @param size the new size of the object in bytes
* @returns 0 on success, negative error code on failure
* Get the value of an extended attribute on a striped object.
*
* @param striper the striper in which the getxattr will occur
- * @param o name of the striped object
+ * @param oid name of the striped object
* @param name which extended attribute to read
* @param buf where to store the result
* @param len size of buf in bytes
* Set an extended attribute on a striped object.
*
* @param striper the striper in which the setxattr will occur
- * @param o name of the object
+ * @param oid name of the object
* @param name which extended attribute to set
* @param buf what to store in the xattr
* @param len the number of bytes in buf
* Always returns 0, as there is no condition for failure.
*
* @param [in] str plain text key/value pairs
+ * @param [in] delims field delimiters to be used for parsing str
* @param [out] str_map key/value pairs parsed from str
- * @param [in] delim field delimiters to be used for parsing str
* @return **0**
*/
extern int get_str_map(
*
* @param results A freshly-created validated_data struct, with values set
* as described in the struct documentation.
- * @param Context The callback to activate once the validation has
- * been completed.
+ * @param mdr The request to be responeded upon the completion of the
+ * validation.
*/
void validate_disk_state(validated_data *results,
MDRequestRef& mdr);
*
* @remarks This function assumes as a default firing value the duration of
* the monitor's lease interval, and adds to it the value specified
- * in @plus
+ * in @e plus
*
* @post expire_event is set
*
set<string> outside_quorum;
/**
- * @defgroup scrub
+ * @defgroup Monitor_h_scrub
* @{
*/
version_t scrub_version; ///< paxos version we are scrubbing
void scrub_reset();
/**
- * @defgroup Synchronization
+ * @defgroup Monitor_h_sync Synchronization
* @{
*/
/**
* Put a whole set of values efficiently and safely.
*
* @param a - prefix/directory
- * @param vals - map of int name -> values
+ * @param start
+ * @param end defines a range of map from int -> values
* @return 0 for success or negative error code
*/
void put_bl_sn_map(const char *a,
* @param old_pool_sum Previous stats sum
* @param last_ts Last timestamp for pool
* @param result_pool_sum Resulting stats
+ * @param result_pool_delta Resulting pool delta
* @param result_ts_delta Resulting timestamp delta
* @param delta_avg_list List of last N computed deltas, used to average
*/
*
* @invariant The received message is an operation of type OP_LEASE
*
- * @param The message sent by the Leader to the Peon during the
+ * @param lease The message sent by the Leader to the Peon during the
* Paxos::extend_lease function
*/
void handle_lease(MMonPaxos *lease);
public:
/**
* @param m A monitor
- * @param mid A machine id
+ * @param name A name for the paxos service. It serves as the naming space
+ * of the underlying persistent storage for this service.
*/
Paxos(Monitor *m, const string &name)
: mon(m),
* Check if a given version is readable.
*
* A version may not be readable for a myriad of reasons:
- * @li the version @v is higher that the last committed version
+ * @li the version @e v is higher that the last committed version
* @li we are not the Leader nor a Peon (election may be on-going)
* @li we do not have a committed value yet
* @li we do not have a valid lease
*/
bool is_readable(version_t seen=0);
/**
- * Read version @v and store its value in @bl
+ * Read version @e v and store its value in @e bl
*
* @param[in] v The version we want to read
* @param[out] bl The version's value
/**
* @param mn A Monitor instance
* @param p A Paxos instance
- * @parem name Our service's name.
+ * @param name Our service's name.
*/
PaxosService(Monitor *mn, Paxos *p, string name)
: mon(mn), paxos(p), service_name(name),
}
/**
- * @defgroup PaxosService_h_Trim
+ * @defgroup PaxosService_h_Trim Functions for trimming states
* @{
*/
/**
void maybe_trim();
/**
- * Auxiliary function to trim our state from version @from to version @to,
- * not including; i.e., the interval [from, to[
+ * Auxiliary function to trim our state from version @p from to version
+ * @p to, not including; i.e., the interval [from, to[
*
* @param t The transaction to which we will add the trim operations.
* @param from the lower limit of the interval to be trimmed
*/
virtual Policy get_default_policy() = 0;
/**
- * Set a Throttler which is applied to all Messages from the given
- * type of peer.
+ * Set Throttlers applied to all Messages from the given type of peer
*
* This is an init-time function and cannot be called after calling
* start() or bind().
*
- * @param type The peer type this Throttler will apply to.
- * @param t The Throttler to apply. The Messenger does not take
- * ownership of this pointer, but you must not destroy it before
- * you destroy the Messenger.
+ * @param type The peer type the Throttlers will apply to.
+ * @param bytes The Throttle for the number of bytes carried by the message
+ * @param msgs The Throttle for the number of messages for this @p type
+ * @note The Messenger does not take ownership of the Throttle pointers, but
+ * you must not destroy them before you destroy the Messenger.
*/
virtual void set_policy_throttlers(int type, Throttle *bytes, Throttle *msgs=NULL) = 0;
/**
*
* @param addr The address of the entity to connect to.
* @param type The peer type of the entity at the address.
- * @param con An existing Connection to associate with the new connection. If
- * NULL, it creates a new Connection.
- * @param msg an initial message to queue on the new connection
*
* @return a pointer to the newly-created connection. Caller does not own a
* reference; take one if you need it.
*
* @param m The Message to queue up. This function eats a reference.
* @param con The existing Connection to use, or NULL if you don't know of one.
- * @param addr The address to send the Message to.
+ * @param dest_addr The address to send the Message to.
* @param dest_type The peer type of the address we're sending to
* just drop silently under failure.
*/
* @param type The peer type of the entity at the address.
* @param con An existing Connection to associate with the new Pipe. If
* NULL, it creates a new Connection.
- * @param msg an initial message to queue on the new pipe
+ * @param first an initial message to queue on the new pipe
*
* @return a pointer to the newly-created Pipe. Caller does not own a
* reference; take one if you need it.
) = 0;
/**
- * Moves objects matching <match> in the lsb <bits>
+ * Moves objects matching @e match in the lsb @e bits
*
* dest and this must be the same subclass
*
const string& type,
const string& data,
const string& journal,
- osflagbits_t flag = 0);
+ osflagbits_t flags = 0);
Logger *logger;
* collection_getattrs - get all xattrs of a collection
*
* @param cid collection name
- * @param asert map of keys and buffers that contain the values
+ * @param aset map of keys and buffers that contain the values
* @returns 0 on success, negative error code on failure
*/
virtual int collection_getattrs(coll_t cid, map<string,bufferptr> &aset)
* @param c collection
* @param start list object that sort >= this value
* @param end list objects that sort < this value
- * @param snapid return no objects with snap < snapid
+ * @param seq return no objects with snap < seq
* @param ls [out] result
* @return zero on success, or negative error
*/
/**
* check if given request parameters match our constraints
*
- * @param auid requesting user's auid
* @param pool_name pool name
* @param nspace_name namespace name
* @param pool_auid pool's auid
*
* @param other pg_log_t to copy from
* @param from copy entries after this version
- * @parem to up to and including this version
+ * @param to up to and including this version
*/
void copy_range(const pg_log_t &other, eversion_t from, eversion_t to);
/**
* copy up to N entries
*
- * @param o source log
+ * @param other source log
* @param max max number of entreis to copy
*/
void copy_up_to(const pg_log_t &other, int max);
* buffer, including the sobject_t soid.
*
* @param ops [in] vector of OSDOps
- * @param in [out] combined data buffer
+ * @param out [out] combined data buffer
*/
static void merge_osd_op_vector_in_data(vector<OSDOp>& ops, bufferlist& out);
* merge outdata members of a vector of OSDOps into a single bufferlist
*
* @param ops [in] vector of OSDOps
- * @param in [out] combined data buffer
+ * @param out [out] combined data buffer
*/
static void merge_osd_op_vector_out_data(vector<OSDOp>& ops, bufferlist& out);
};
virtual void ms_handle_remote_reset(Connection *con);
/**
- * @defgroup Authentication
+ * @defgroup test_xio_dispatcher_h_auth Authentication
* @{
*/
/**
projects / ceph.git / commitdiff