Remove cassandra queries to user_keys_hash

they are never read anymore since 'onboarding' / auto-connect was removed in #1005

changelog.d/5-internal/remove-hashed-key-queries

@@ -0,0 +1 @@
+Remove cassandra queries to the user_keys_hash table, as they are never read anymore since 'onboarding' / auto-connect was removed in https://github.com/wireapp/wire-server/pull/1005

services/brig/schema/src/Main.hs

@@ -161,5 +161,7 @@ main = do
 
       -- FUTUREWORK: undo V41 (searchable flag); we stopped using it in
       -- https://github.com/wireapp/wire-server/pull/964
+      --
+      -- FUTUREWORK after July 2023: integrate V_FUTUREWORK here.
     ]
     `finally` Log.close l

services/brig/schema/src/V_FUTUREWORK.hs

@@ -0,0 +1,48 @@
+{-# LANGUAGE QuasiQuotes #-}
+
+-- This file is part of the Wire Server implementation.
+--
+-- Copyright (C) 2022 Wire Swiss GmbH <opensource@wire.com>
+--
+-- This program is free software: you can redistribute it and/or modify it under
+-- the terms of the GNU Affero General Public License as published by the Free
+-- Software Foundation, either version 3 of the License, or (at your option) any
+-- later version.
+--
+-- This program is distributed in the hope that it will be useful, but WITHOUT
+-- ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or FITNESS
+-- FOR A PARTICULAR PURPOSE. See the GNU Affero General Public License for more
+-- details.
+--
+-- You should have received a copy of the GNU Affero General Public License along
+-- with this program. If not, see <https://www.gnu.org/licenses/>.
+
+module V_FUTUREWORK
+  ( migration,
+  )
+where
+
+import Cassandra.Schema
+import Imports
+import Text.RawString.QQ
+
+-- user_keys_hash usage was removed in https://github.com/wireapp/wire-server/pull/2902
+--
+-- However, it's dangerous to remove a cassandra table together with the usage, as
+-- during deployment, there is a time window where the schema migration has run, but the
+-- old code still serves traffic, which then leads to 5xxs and user-observable errors.
+-- Therefore the policy is to wait a reasonable amount of time (6 months) to allow all
+-- installations to upgrade before removing the database tables. See also
+-- backwards-incompatbile schema migration docs in
+-- https://docs.wire.com/developer/developer/cassandra-interaction.html?highlight=backwards+incompatbile#backwards-incompatible-schema-changes
+--
+-- FUTUREWORK: uncomment the code below after July 2023, rename this module with a version number, and
+-- integrate it inside Main.hs and App.hs
+migration :: Migration
+migration = undefined
+
+-- Migration FUTUREWORK_NUMBER "Drop deprecated user_keys_hashed table" $ do
+--   schema'
+--     [r|
+--       DROP TABLE IF EXISTS user_keys_hash
+--    |]

services/brig/src/Brig/Data/UserKey.hs

@@ -30,24 +30,17 @@ module Brig.Data.UserKey
     lookupKey,
     deleteKey,
     deleteKeyForUser,
-    lookupPhoneHashes,
   )
 where
 
-import Brig.App (Env, digestSHA256)
+import Brig.App (Env)
 import Brig.Data.Instances ()
 import qualified Brig.Data.User as User
 import Brig.Email
 import Brig.Phone
 import Cassandra
-import Control.Lens (view)
-import qualified Data.ByteString as B
-import Data.ByteString.Lazy (toStrict)
 import Data.Id
-import qualified Data.Multihash.Digest as MH
-import qualified Data.Text.Encoding as T
 import Imports
-import OpenSSL.EVP.Digest (digestBS)
 import Wire.API.User (fromEmail)
 
 -- | A natural identifier (i.e. unique key) of a user.
@@ -60,35 +53,6 @@ instance Eq UserKey where
   (UserPhoneKey k) == (UserPhoneKey k') = k == k'
   _ == _ = False
 
-data UKHashType
-  = UKHashPhone
-  | UKHashEmail
-  deriving (Eq)
-
-instance Cql UKHashType where
-  ctype = Tagged IntColumn
-
-  fromCql (CqlInt i) = case i of
-    0 -> pure UKHashPhone
-    1 -> pure UKHashEmail
-    n -> Left $ "unexpected hashtype: " ++ show n
-  fromCql _ = Left "userkeyhashtype: int expected"
-
-  toCql UKHashPhone = CqlInt 0
-  toCql UKHashEmail = CqlInt 1
-
-newtype UserKeyHash = UserKeyHash MH.MultihashDigest
-
-instance Cql UserKeyHash where
-  ctype = Tagged BlobColumn
-
-  fromCql (CqlBlob lbs) = case MH.decode (toStrict lbs) of
-    Left e -> Left ("userkeyhash: " ++ e)
-    Right h -> pure $ UserKeyHash h
-  fromCql _ = Left "userkeyhash: expected blob"
-
-  toCql (UserKeyHash d) = CqlBlob $ MH.encode (MH.algorithm d) (MH.digest d)
-
 userEmailKey :: Email -> UserKey
 userEmailKey = UserEmailKey . mkEmailKey
 
@@ -154,15 +118,10 @@ lookupKey k =
 
 insertKey :: (MonadClient m, MonadReader Env m) => UserId -> UserKey -> m ()
 insertKey u k = do
-  hk <- hashKey k
-  let kt = foldKey (\(_ :: Email) -> UKHashEmail) (\(_ :: Phone) -> UKHashPhone) k
-  retry x5 $ write insertHashed (params LocalQuorum (hk, kt, u))
   retry x5 $ write keyInsert (params LocalQuorum (keyText k, u))
 
 deleteKey :: (MonadClient m, MonadReader Env m) => UserKey -> m ()
 deleteKey k = do
-  hk <- hashKey k
-  retry x5 $ write deleteHashed (params LocalQuorum (Identity hk))
   retry x5 $ write keyDelete (params LocalQuorum (Identity $ keyText k))
 
 -- | Delete `UserKey` for `UserId`
@@ -180,21 +139,6 @@ deleteKeyForUser uid k = do
     Just keyUid | keyUid == uid -> deleteKey k
     _ -> pure ()
 
-hashKey :: MonadReader Env m => UserKey -> m UserKeyHash
-hashKey uk = do
-  d <- view digestSHA256
-  let d' = digestBS d $ T.encodeUtf8 (keyText uk)
-  pure . UserKeyHash $
-    MH.MultihashDigest MH.SHA256 (B.length d') d'
-
-lookupPhoneHashes :: MonadClient m => [ByteString] -> m [(ByteString, UserId)]
-lookupPhoneHashes hp =
-  mapMaybe mk <$> retry x1 (query selectHashed (params One (Identity hashed)))
-  where
-    hashed = fmap (\h -> UserKeyHash $ MH.MultihashDigest MH.SHA256 (B.length h) h) hp
-    mk (UserKeyHash d, UKHashPhone, u) = Just (MH.digest d, u)
-    mk (_, _, _) = Nothing
-
 --------------------------------------------------------------------------------
 -- Queries
 
@@ -206,12 +150,3 @@ keySelect = "SELECT user FROM user_keys WHERE key = ?"
 
 keyDelete :: PrepQuery W (Identity Text) ()
 keyDelete = "DELETE FROM user_keys WHERE key = ?"
-
-insertHashed :: PrepQuery W (UserKeyHash, UKHashType, UserId) ()
-insertHashed = "INSERT INTO user_keys_hash(key, key_type, user) VALUES (?, ?, ?)"
-
-deleteHashed :: PrepQuery W (Identity UserKeyHash) ()
-deleteHashed = "DELETE FROM user_keys_hash WHERE key = ?"
-
-selectHashed :: PrepQuery R (Identity [UserKeyHash]) (UserKeyHash, UKHashType, UserId)
-selectHashed = "SELECT key, key_type, user FROM user_keys_hash WHERE key IN ?"

tools/db/move-team/src/ParseSchema.hs

@@ -377,11 +377,6 @@ main = do
                        --   PRIMARY KEY (key :: text) (Brig.Data.UserKey)
                        --   FUTUREWORK: do we need it?  can we do better than a full table scan?
                        mkChunk' "brig" "user_keys" "[Int32]" "keys" "key in ?",
-                       -- brig.user_keys_hash
-                       --   FUTUREWORK: do we need it?
-                       --   can we do better than a full table scan?
-                       --   mkChunk' "brig" "user_keys_hash" "[UserKeyHash]" "keys" "key in ?"
-                       mkChunk' "brig" "user_keys_hash" "[Int32]" "keys" "key in ?",
                        -- brig.vcodes
                        -- galley.billing_team_member
                        --   PRIMARY KEY (team, user)

tools/db/move-team/src/Schema.hs

@@ -496,52 +496,6 @@ importBrigUserKeys Env {..} path = do
       putStrLn $ "Skipping because not found: " <> path
       pure ()
 
--- brig.user_keys_hash
-
-type RowBrigUserKeysHash = (Maybe Blob, Maybe Int32, Maybe UUID)
-
-selectBrigUserKeysHash :: PrepQuery R (Identity [Int32]) RowBrigUserKeysHash
-selectBrigUserKeysHash = "SELECT key, key_type, user FROM user_keys_hash WHERE key in ?"
-
-readBrigUserKeysHash :: Env -> [Int32] -> ConduitM () [RowBrigUserKeysHash] IO ()
-readBrigUserKeysHash Env {..} keys =
-  transPipe (runClient envBrig) $
-    paginateC selectBrigUserKeysHash (paramsP LocalQuorum (pure keys) envPageSize) x5
-
-selectBrigUserKeysHashAll :: PrepQuery R () RowBrigUserKeysHash
-selectBrigUserKeysHashAll = "SELECT key, key_type, user FROM user_keys_hash"
-
-readBrigUserKeysHashAll :: Env -> ConduitM () [RowBrigUserKeysHash] IO ()
-readBrigUserKeysHashAll Env {..} =
-  transPipe (runClient envBrig) $
-    paginateC selectBrigUserKeysHashAll (paramsP LocalQuorum () envPageSize) x5
-
-exportBrigUserKeysHashFull :: Env -> FilePath -> IO ()
-exportBrigUserKeysHashFull env path = do
-  putStrLn $ "Exporting " <> "brig.user_keys_hash" <> " to " <> path
-  withBinaryFile path WriteMode $ \handle ->
-    runConduit $
-      readBrigUserKeysHashAll env
-        .| sinkJsonLines handle
-
-insertBrigUserKeysHash :: PrepQuery W RowBrigUserKeysHash ()
-insertBrigUserKeysHash =
-  "INSERT INTO user_keys_hash (key, key_type, user) VALUES (?, ?, ?)"
-
-importBrigUserKeysHash :: Env -> FilePath -> IO ()
-importBrigUserKeysHash Env {..} path = do
-  exists <- doesFileExist path
-  if exists
-    then do
-      putStrLn $ "Importing " <> path <> " to " <> "brig.user_keys_hash"
-      withBinaryFile path ReadMode $ \handle ->
-        runConduit $
-          sourceJsonLines handle
-            .| transPipe (runClient envBrig) (sinkTableRows insertBrigUserKeysHash)
-    else do
-      putStrLn $ "Skipping because not found: " <> path
-      pure ()
-
 -- galley.billing_team_member
 
 type RowGalleyBillingTeamMember = (Maybe UUID, Maybe UUID)
@@ -1244,7 +1198,6 @@ importAllTables env@Env {..} = do
   importBrigUser env (envTargetPath </> "brig.user")
   importBrigUserHandle env (envTargetPath </> "brig.user_handle")
   importBrigUserKeys env (envTargetPath </> "brig.user_keys")
-  importBrigUserKeysHash env (envTargetPath </> "brig.user_keys_hash")
   importGalleyBillingTeamMember env (envTargetPath </> "galley.billing_team_member")
   importGalleyClients env (envTargetPath </> "galley.clients")
   importGalleyConversation env (envTargetPath </> "galley.conversation")
@@ -1273,7 +1226,6 @@ exportAllTablesFull env@Env {..} = do
   exportBrigUserFull env (envTargetPath </> "brig.user")
   exportBrigUserHandleFull env (envTargetPath </> "brig.user_handle")
   exportBrigUserKeysFull env (envTargetPath </> "brig.user_keys")
-  exportBrigUserKeysHashFull env (envTargetPath </> "brig.user_keys_hash")
   exportGalleyBillingTeamMemberFull env (envTargetPath </> "galley.billing_team_member")
   exportGalleyClientsFull env (envTargetPath </> "galley.clients")
   exportGalleyConversationFull env (envTargetPath </> "galley.conversation")

tools/db/move-team/src/Work.hs

@@ -157,8 +157,6 @@ runFullScans env@Env {..} users = do
       .| mapC (filter (haveId . view _2))
 
   -- FUTUREWORK: no need to read this table, it can be populated from `brig.user`
-  appendJsonLines (envTargetPath </> "brig.user_keys_hash") $
-    readBrigUserKeysHashAll env
   appendJsonLines (envTargetPath </> "spar.user") $
     readSparUserAll env
       .| mapC (filter (haveId . view _3))