close #29: periodic cache maintenance

periodically delete expired cache entries, check unverified ones and
potentially use them as neighbour

src/Hash2Pub/DHTProtocol.hs

@@ -20,11 +20,16 @@ module Hash2Pub.DHTProtocol
     , requestPing
     , requestStabilise
     , queryIdLookupLoop
+    , queueAddEntries
+    , queueDeleteEntries
+    , queueDeleteEntry
     , resolve
     , mkSendSocket
     , mkServerSocket
     , handleIncomingRequest
     , ackRequest
+    , isPossibleSuccessor
+    , isPossiblePredecessor
     )
         where
 
@@ -81,7 +86,7 @@ import           Debug.Trace                    (trace)
 queryLocalCache :: LocalNodeState -> NodeCache -> Int -> NodeID -> QueryResponse
 queryLocalCache ownState nCache lBestNodes targetID
     -- as target ID falls between own ID and first predecessor, it is handled by this node
-      | isInOwnResponsibilitySlice ownState targetID = FOUND . toRemoteNodeState $ ownState
+      | targetID `isInOwnResponsibilitySlice` ownState = FOUND . toRemoteNodeState $ ownState
     -- my interpretation: the "l best next hops" are the l-1 closest preceding nodes and
     -- the closest succeeding node (like with the p initiated parallel queries
       | otherwise = FORWARD $ closestSuccessor `Set.union` closestPredecessors
@@ -110,8 +115,8 @@ queryLocalCache ownState nCache lBestNodes targetID
 -- Looks up the successor of the lookup key on a 'RingMap' representation of the
 -- predecessor list with the node itself added. If the result is the same as the node
 -- itself then it falls into the responsibility interval.
-isInOwnResponsibilitySlice :: HasKeyID a => LocalNodeState -> a -> Bool
-isInOwnResponsibilitySlice ownNs lookupTarget = (getKeyID <$> rMapLookupSucc (getKeyID lookupTarget) predecessorRMap) == pure (getNid ownNs)
+isInOwnResponsibilitySlice :: HasKeyID a => a -> LocalNodeState -> Bool
+isInOwnResponsibilitySlice lookupTarget ownNs = (getKeyID <$> rMapLookupSucc (getKeyID lookupTarget) predecessorRMap) == pure (getNid ownNs)
   where
     predecessorList = predecessors ownNs
     -- add node itself to RingMap representation, to distinguish between
@@ -119,6 +124,16 @@ isInOwnResponsibilitySlice ownNs lookupTarget = (getKeyID <$> rMapLookupSucc (ge
     predecessorRMap = addRMapEntry (toRemoteNodeState ownNs) $ rMapFromList predecessorList
     closestPredecessor = headMay predecessorList
 
+isPossiblePredecessor :: HasKeyID a => a -> LocalNodeState -> Bool
+isPossiblePredecessor = isInOwnResponsibilitySlice
+
+isPossibleSuccessor :: HasKeyID a => a -> LocalNodeState -> Bool
+isPossibleSuccessor lookupTarget ownNs = (getKeyID <$> rMapLookupPred (getKeyID lookupTarget) successorRMap) == pure (getNid ownNs)
+  where
+    successorList = successors ownNs
+    successorRMap = addRMapEntry (toRemoteNodeState ownNs) $ rMapFromList successorList
+    closestSuccessor = headMay successorList
+
 -- cache operations
 
 -- | update or insert a 'RemoteCacheEntry' into the cache,
@@ -662,6 +677,21 @@ queueAddEntries entries ns = do
     now <- getPOSIXTime
     forM_ entries $ \entry -> atomically $ writeTQueue (cacheWriteQueue ns)  $ addCacheEntryPure now entry
 
+
+-- | enque a list of node IDs to be deleted from the global NodeCache
+queueDeleteEntries :: Foldable c
+                   => c NodeID
+                   -> LocalNodeState
+                   -> IO ()
+queueDeleteEntries ids ns = forM_ ids $ atomically . writeTQueue (cacheWriteQueue ns) . deleteCacheEntry
+
+
+-- | enque a single node ID to be deleted from the global NodeCache
+queueDeleteEntry :: NodeID
+                 -> LocalNodeState
+                 -> IO ()
+queueDeleteEntry toDelete = queueDeleteEntries $ Identity toDelete
+
 -- | retry an IO action at most *i* times until it delivers a result
 attempts :: Int             -- ^ number of retries *i*
          -> IO (Maybe a)    -- ^ action to retry

src/Hash2Pub/FediChord.hs

@@ -63,6 +63,7 @@ import qualified Data.ByteString               as BS
 import qualified Data.ByteString.UTF8          as BSU
 import           Data.Either                   (rights)
 import           Data.Foldable                 (foldr')
+import           Data.Functor.Identity
 import           Data.IP                       (IPv6, fromHostAddress6,
                                                 toHostAddress6)
 import           Data.List                     ((\\))
@@ -183,6 +184,58 @@ cacheWriter nsSTM =
         modifyTVar' (nodeCacheSTM ns) cacheModifier
 
 
+-- TODO: make max entry age configurable
+maxEntryAge :: POSIXTime
+maxEntryAge = 600
+
+
+-- | Periodically iterate through cache, clean up expired entries and verify unverified ones
+cacheVerifyThread :: LocalNodeStateSTM -> IO ()
+cacheVerifyThread nsSTM = forever $ do
+    -- get cache
+    (ns, cache) <- atomically $ do
+        ns <- readTVar nsSTM
+        cache <- readTVar $ nodeCacheSTM ns
+        pure (ns, cache)
+    -- iterate entries:
+    -- for avoiding too many time syscalls, get current time before iterating.
+    now <- getPOSIXTime
+    forM_ (cacheEntries cache) (\(CacheEntry validated node ts) ->
+        -- case too old: delete (future work: decide whether pinging and resetting timestamp is better)
+        if (now - ts) > maxEntryAge
+           then
+           queueDeleteEntry (getNid node) ns
+    -- case unverified: try verifying, otherwise delete
+           else if not validated
+                then do
+                    -- marking as verified is done by 'requestPing' as well
+                    pong <- requestPing ns node
+                    either (\_->
+                        queueDeleteEntry (getNid node) ns
+                           )
+                           (\vss ->
+                               if node `notElem` vss
+                                  then queueDeleteEntry (getNid node) ns
+                                 -- after verifying a node, check whether it can be a closer neighbour
+                                 else do
+                                     if node `isPossiblePredecessor` ns
+                                        then atomically $ do
+                                            ns' <- readTVar nsSTM
+                                            writeTVar nsSTM $ addPredecessors [node] ns'
+                                        else pure ()
+                                     if node `isPossibleSuccessor` ns
+                                        then atomically $ do
+                                            ns' <- readTVar nsSTM
+                                            writeTVar nsSTM $ addSuccessors [node] ns'
+                                        else pure ()
+                           ) pong
+           else pure ()
+                               )
+
+    threadDelay $ toEnum (fromEnum maxEntryAge `div` 20)
+
+
+
 -- | Periodically send @StabiliseRequest' s to the closest neighbour nodes, until
 -- one responds, and get their neighbours for maintaining the own neighbour lists.
 -- If necessary, request new neighbours.
@@ -308,9 +361,10 @@ fediMainThreads sock nsSTM = do
     concurrently_
         (fediMessageHandler sendQ recvQ nsSTM) $
         concurrently_ (stabiliseThread nsSTM) $
-            concurrently_
-                (sendThread sock sendQ)
-                (recvThread sock recvQ)
+            concurrently_ (cacheVerifyThread nsSTM) $
+                concurrently_
+                    (sendThread sock sendQ)
+                    (recvThread sock recvQ)
 
 
 -- defining this here as, for now, the RequestMap is only used by fediMessageHandler.
@@ -322,17 +376,17 @@ data RequestMapEntry = RequestMapEntry (Set.Set FediChordMessage) (Maybe Integer
 
 -- TODO: make purge age configurable
 -- | periodically clean up old request parts
-purgeAge :: POSIXTime
-purgeAge = 60 -- seconds
+responsePurgeAge :: POSIXTime
+responsePurgeAge = 60 -- seconds
 
 requestMapPurge :: MVar RequestMap -> IO ()
 requestMapPurge mapVar = forever $ do
     rMapState <- takeMVar mapVar
     now <- getPOSIXTime
     putMVar mapVar $ Map.filter (\entry@(RequestMapEntry _ _ ts)  ->
-        now - ts < purgeAge
+        now - ts < responsePurgeAge
                                 ) rMapState
-    threadDelay $ fromEnum purgeAge * 2000
+    threadDelay $ fromEnum responsePurgeAge * 2000
 
 
 -- | Wait for messages, deserialise them, manage parts and acknowledgement status,