freebsd-skq/usr.sbin/nscd/cachelib.h
Stefan Eßer a397989d4e Add the possibility to specify a threshold for the number of negative cache
results required to have the cache return lookup failure.

A new configuration parameter is introduced, which must be set to a value
greater than 1 to activate this feature. The default behavior is unchanged.

The purpose of this change is to allow probes for the existence of an entry
(which are expected to fail), before that entry is added to one of the
queried databases, without the cache returning the stale information from
the probe query until that cache entry expires. If, for example, a new user
account is created after checking that the new account name is available,
the negative cache entry would prevent immediate access to the account.

For that example, the new configuration option

negative-confidence-threshold passwd 2

will require a second negative query result to consider the negative cache
entry for a passwd entry valid, but if the user account has been created
between the queries, then the positive query result from the second query
will be cached and returned.
2012-07-04 09:02:12 +00:00

264 lines
8.0 KiB
C

/*-
* Copyright (c) 2005 Michael Bushkov <bushman@rsu.ru>
* All rights reserved.
*
* Redistribution and use in source and binary forms, with or without
* modification, are permitted provided that the following conditions
* are met:
* 1. Redistributions of source code must retain the above copyright
* notice, this list of conditions and the following disclaimer.
* 2. Redistributions in binary form must reproduce the above copyright
* notice, this list of conditions and the following disclaimer in the
* documentation and/or other materials provided with the distribution.
*
* THIS SOFTWARE IS PROVIDED BY THE AUTHOR AND CONTRIBUTORS ``AS IS'' AND
* ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
* IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
* ARE DISCLAIMED. IN NO EVENT SHALL THE AUTHOR OR CONTRIBUTORS BE LIABLE
* FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL
* DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS
* OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION)
* HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT
* LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY
* OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF
* SUCH DAMAGE.
*
* $FreeBSD$
*/
#ifndef __NSCD_CACHELIB_H__
#define __NSCD_CACHELIB_H__
#include "hashtable.h"
#include "cacheplcs.h"
enum cache_entry_t {
CET_COMMON = 0, /* cache item is atomic */
CET_MULTIPART /* cache item is formed part by part */
};
enum cache_transformation_t {
CTT_FLUSH = 0, /* flush the cache - delete all obsolete items */
CTT_CLEAR = 1 /* delete all items in the cache */
};
/* cache deletion policy type enum */
enum cache_policy_t {
CPT_FIFO = 0, /* first-in first-out */
CPT_LRU = 1, /* least recently used */
CPT_LFU = 2 /* least frequently used */
};
/* multipart sessions can be used for reading and writing */
enum cache_mp_session_t {
CMPT_READ_SESSION,
CMPT_WRITE_SESSION
};
/*
* When doing partial transformations of entries (which are applied for
* elements with keys, that contain specified buffer in its left or
* right part), this enum will show the needed position of the key part.
*/
enum part_position_t {
KPPT_LEFT,
KPPT_RIGHT
};
/* num_levels attribute is obsolete, i think - user can always emulate it
* by using one entry.
* get_time_func is needed to have the clocks-independent counter
*/
struct cache_params {
void (*get_time_func)(struct timeval *);
};
/*
* base structure - normal_cache_entry_params and multipart_cache_entry_params
* are "inherited" from it
*/
struct cache_entry_params {
enum cache_entry_t entry_type;
char *entry_name;
};
/* params, used for most entries */
struct common_cache_entry_params {
struct cache_entry_params cep;
size_t cache_entries_size;
size_t max_elemsize; /* if 0 then no check is made */
size_t satisf_elemsize; /* if entry size is exceeded,
* this number of elements will be left,
* others will be deleted */
int confidence_threshold; /* number matching replies required */
struct timeval max_lifetime; /* if 0 then no check is made */
enum cache_policy_t policy; /* policy used for transformations */
};
/* params, used for multipart entries */
struct mp_cache_entry_params {
struct cache_entry_params cep;
/* unique fields */
size_t max_elemsize; /* if 0 then no check is made */
size_t max_sessions; /* maximum number of active sessions */
struct timeval max_lifetime; /* maximum elements lifetime */
};
struct cache_ht_item_data_ {
/* key is the bytes sequence only - not the null-terminated string */
char *key;
size_t key_size;
char *value;
size_t value_size;
struct cache_policy_item_ *fifo_policy_item;
int confidence; /* incremented for each verification */
};
struct cache_ht_item_ {
HASHTABLE_ENTRY_HEAD(ht_item_, struct cache_ht_item_data_) data;
};
struct cache_entry_ {
char *name;
struct cache_entry_params *params;
};
struct cache_common_entry_ {
char *name;
struct cache_entry_params *params;
struct common_cache_entry_params common_params;
HASHTABLE_HEAD(cache_ht_, cache_ht_item_) items;
size_t items_size;
/*
* Entry always has the FIFO policy, that is used to eliminate old
* elements (the ones, with lifetime more than max_lifetime). Besides,
* user can specify another policy to be applied, when there are too
* many elements in the entry. So policies_size can be 1 or 2.
*/
struct cache_policy_ **policies;
size_t policies_size;
void (*get_time_func)(struct timeval *);
};
struct cache_mp_data_item_ {
char *value;
size_t value_size;
TAILQ_ENTRY(cache_mp_data_item_) entries;
};
struct cache_mp_write_session_ {
struct cache_mp_entry_ *parent_entry;
/*
* All items are accumulated in this queue. When the session is
* committed, they all will be copied to the multipart entry.
*/
TAILQ_HEAD(cache_mp_data_item_head, cache_mp_data_item_) items;
size_t items_size;
TAILQ_ENTRY(cache_mp_write_session_) entries;
};
struct cache_mp_read_session_ {
struct cache_mp_entry_ *parent_entry;
struct cache_mp_data_item_ *current_item;
TAILQ_ENTRY(cache_mp_read_session_) entries;
};
struct cache_mp_entry_ {
char *name;
struct cache_entry_params *params;
struct mp_cache_entry_params mp_params;
/* All opened write sessions */
TAILQ_HEAD(write_sessions_head, cache_mp_write_session_) ws_head;
size_t ws_size;
/* All opened read sessions */
TAILQ_HEAD(read_sessions_head, cache_mp_read_session_) rs_head;
size_t rs_size;
/*
* completed_write_session is the committed write sessions. All read
* sessions use data from it. If the completed_write_session is out of
* date, but still in use by some of the read sessions, the newly
* committed write session is stored in the pending_write_session.
* In such a case, completed_write_session will be substituted with
* pending_write_session as soon as it won't be used by any of
* the read sessions.
*/
struct cache_mp_write_session_ *completed_write_session;
struct cache_mp_write_session_ *pending_write_session;
struct timeval creation_time;
struct timeval last_request_time;
void (*get_time_func)(struct timeval *);
};
struct cache_ {
struct cache_params params;
struct cache_entry_ **entries;
size_t entries_capacity;
size_t entries_size;
};
/* simple abstractions - for not to write "struct" every time */
typedef struct cache_ *cache;
typedef struct cache_entry_ *cache_entry;
typedef struct cache_mp_write_session_ *cache_mp_write_session;
typedef struct cache_mp_read_session_ *cache_mp_read_session;
#define INVALID_CACHE (NULL)
#define INVALID_CACHE_ENTRY (NULL)
#define INVALID_CACHE_MP_WRITE_SESSION (NULL)
#define INVALID_CACHE_MP_READ_SESSION (NULL)
/*
* NOTE: all cache operations are thread-unsafe. You must ensure thread-safety
* externally, by yourself.
*/
/* cache initialization/destruction routines */
cache init_cache(struct cache_params const *);
void destroy_cache(cache);
/* cache entries manipulation routines */
int register_cache_entry(cache, struct cache_entry_params const *);
int unregister_cache_entry(cache, const char *);
cache_entry find_cache_entry(cache, const char *);
/* read/write operations used on common entries */
int cache_read(cache_entry, const char *, size_t, char *, size_t *);
int cache_write(cache_entry, const char *, size_t, char const *, size_t);
/* read/write operations used on multipart entries */
cache_mp_write_session open_cache_mp_write_session(cache_entry);
int cache_mp_write(cache_mp_write_session, char *, size_t);
void abandon_cache_mp_write_session(cache_mp_write_session);
void close_cache_mp_write_session(cache_mp_write_session);
cache_mp_read_session open_cache_mp_read_session(cache_entry);
int cache_mp_read(cache_mp_read_session, char *, size_t *);
void close_cache_mp_read_session(cache_mp_read_session);
/* transformation routines */
int transform_cache_entry(cache_entry, enum cache_transformation_t);
int transform_cache_entry_part(cache_entry, enum cache_transformation_t,
const char *, size_t, enum part_position_t);
#endif