Add support for the EntryCacheNoWaitRefreshTimeout

This timeout specifies the lifetime of a cache entry before it is
updated out-of-band. When this timeout is hit, the request will
still complete from cache, but the SSSD will also go and update
the cached entry in the background to extend the life of the
cache entry and reduce the wait time of a future request.

5 files changed, 75 insertions, 2 deletions

diff --git a/server/examples/sssd.conf b/server/examples/sssd.conf
index 90e0c8d1..b47ab9dd 100644
--- a/server/examples/sssd.conf
+++ b/server/examples/sssd.conf
@@ -13,6 +13,15 @@ description = NSS Responder Configuration
 filterGroups = root
 filterUsers = root
 
+# The EntryCacheTimeout indicates the number of seconds to retain before
+# an entry in cache is considered stale and must block to refresh.
+# The EntryCacheNoWaitRefreshTimeout indicates the number of seconds to
+# wait before updating the cache out-of-band. (NSS requests will still
+# be returned from cache until the full EntryCacheTimeout). Setting this
+# value to 0 turns this feature off (default)
+; EntryCacheTimeout = 600
+; EntryCacheNoWaitRefreshTimeout = 300
+
 [services/dp]
 description = Data Provider Configuration
 
diff --git a/server/man/sssd.conf.5.xml b/server/man/sssd.conf.5.xml
index b64899a0..827dd96b 100644
--- a/server/man/sssd.conf.5.xml
+++ b/server/man/sssd.conf.5.xml
@@ -336,6 +336,19 @@
                     </listitem>
                 </varlistentry>
                 <varlistentry>
+                    <term>EntryCacheNoWaitRefreshTimeout (integer)</term>
+                    <listitem>
+                        <para>
+                            How long should nss_sss return cached entries before
+                            initiating an out-of-band cache refresh (0 disables
+                            this feature)
+                        </para>
+                        <para>
+                            Default: 0
+                        </para>
+                    </listitem>
+                </varlistentry>
+                <varlistentry>
                     <term>EntryNegativeTimeout (integer)</term>
                     <listitem>
                         <para>
diff --git a/server/responder/nss/nsssrv.c b/server/responder/nss/nsssrv.c
index ad804389..3920189a 100644
--- a/server/responder/nss/nsssrv.c
+++ b/server/responder/nss/nsssrv.c
@@ -107,6 +107,22 @@ static int nss_get_config(struct nss_ctx *nctx,
                          &nctx->filter_users_in_groups);
     if (ret != EOK) goto done;
 
+
+    ret = confdb_get_int(cdb, nctx, NSS_SRV_CONFIG,
+                         "EntryCacheNoWaitRefreshTimeout", 0,
+                         &nctx->cache_refresh_timeout);
+    if (ret != EOK) goto done;
+    if (nctx->cache_refresh_timeout >= nctx->cache_timeout) {
+        DEBUG(0,("Configuration error: EntryCacheNoWaitRefreshTimeout exceeds"
+                 "EntryCacheTimeout. Disabling feature.\n"));
+        nctx->cache_refresh_timeout = 0;
+    }
+    if (nctx->cache_refresh_timeout < 0) {
+        DEBUG(0,("Configuration error: EntryCacheNoWaitRefreshTimeout is"
+                 "invalid. Disabling feature.\n"));
+        nctx->cache_refresh_timeout = 0;
+    }
+
     ret = confdb_get_string_as_list(cdb, tmpctx, NSS_SRV_CONFIG,
                                     "filterUsers", &filter_list);
     if (ret == ENOENT) filter_list = NULL;
diff --git a/server/responder/nss/nsssrv.h b/server/responder/nss/nsssrv.h
index c5a7bb30..a5adbaf8 100644
--- a/server/responder/nss/nsssrv.h
+++ b/server/responder/nss/nsssrv.h
@@ -46,10 +46,12 @@ struct getent_ctx;
 struct nss_ctx {
     struct resp_ctx *rctx;
 
-    int cache_timeout;
     int neg_timeout;
     struct nss_nc_ctx *ncache;
 
+    int cache_timeout;
+    int cache_refresh_timeout;
+
     int enum_cache_timeout;
     time_t last_user_enum;
     time_t last_group_enum;
diff --git a/server/responder/nss/nsssrv_cmd.c b/server/responder/nss/nsssrv_cmd.c
index 3b0f867e..92600933 100644
--- a/server/responder/nss/nsssrv_cmd.c
+++ b/server/responder/nss/nsssrv_cmd.c
@@ -278,26 +278,50 @@ static errno_t check_cache(struct nss_dom_ctx *dctx,
 {
     errno_t ret;
     int timeout;
+    int refresh_timeout;
     time_t now;
     uint64_t lastUpdate;
     struct nss_cmd_ctx *cmdctx = dctx->cmdctx;
     struct cli_ctx *cctx = cmdctx->cctx;
     bool call_provider = false;
+    sss_dp_callback_t cb = NULL;
 
     if (dctx->check_provider) {
         switch (res->count) {
         case 0:
+            /* This is a cache miss. We need to get the updated user
+             * information before returning it.
+             */
             call_provider = true;
+            cb = callback;
             break;
 
         case 1:
             timeout = nctx->cache_timeout;
+            refresh_timeout = nctx->cache_refresh_timeout;
             now = time(NULL);
 
             lastUpdate = ldb_msg_find_attr_as_uint64(res->msgs[0],
                                                      SYSDB_LAST_UPDATE, 0);
             if (lastUpdate + timeout < now) {
+                /* This is a cache miss. We need to get the updated user
+                 * information before returning it.
+                 */
                 call_provider = true;
+                cb = callback;
+            }
+            else if (refresh_timeout && (lastUpdate + refresh_timeout < now)) {
+                /* We're past the the cache refresh timeout
+                 * We'll return the value from the cache, but we'll also
+                 * queue the cache entry for update out-of-band.
+                 */
+                call_provider = true;
+                cb = NULL;
+            }
+            else {
+                /* Cache is still valid. Just return it. */
+                call_provider = false;
+                cb = NULL;
             }
             break;
 
@@ -340,7 +364,16 @@ static errno_t check_cache(struct nss_dom_ctx *dctx,
         /* Tell the calling function to return so the dp callback
          * can resolve
          */
-        return EAGAIN;
+        if (cb) {
+            return EAGAIN;
+        }
+
+        /* No callback required
+         * This was an out-of-band update. We'll return EOK
+         * so the calling function can return the cached entry
+         * immediately.
+         */
+        DEBUG(3, ("Updating cache out-of-band\n"));
     }
 
     return EOK;