From aecfea27593bc121273fe53a6c11d4a22567004f Mon Sep 17 00:00:00 2001 From: Joey Hess Date: Mon, 14 Aug 2017 13:55:38 -0400 Subject: Added GIT_ANNEX_VECTOR_CLOCK environment variable Can be used to override the default timestamps used in log files in the git-annex branch. This is a dangerous environment variable; use with caution. Note that this only affects writing to the logs on the git-annex branch. It is not used for metadata in git commits (other env vars can be set for that). There are many other places where timestamps are still used, that don't get committed to git, but do touch disk. Including regular timestamps of files, and timestamps embedded in some files in .git/annex/, including the last fsck timestamp and timestamps in transfer log files. A good way to find such things in git-annex is to get for getPOSIXTime and getCurrentTime, although some of the results are of course false positives that never hit disk (unless git-annex gets swapped out..) So this commit does NOT necessarily make git-annex comply with some HIPPA privacy regulations; it's up to the user to determine if they can use it in a way compliant with such regulations. Benchmarking: It takes 0.00114 milliseconds to call getEnv "GIT_ANNEX_VECTOR_CLOCK" when that env var is not set. So, 100 thousand log files can be written with an added overhead of only 0.114 seconds. That should be by far swamped by the actual overhead of writing the log files and making the commit containing them. This commit was supported by the NSF-funded DataLad project. --- Command/Expire.hs | 9 +++++---- Command/Forget.hs | 9 ++++----- Command/MetaData.hs | 28 ++++++++++++++-------------- 3 files changed, 23 insertions(+), 23 deletions(-) (limited to 'Command') diff --git a/Command/Expire.hs b/Command/Expire.hs index 8dd0e962e..551742304 100644 --- a/Command/Expire.hs +++ b/Command/Expire.hs @@ -13,6 +13,7 @@ import Logs.UUID import Logs.MapLog import Logs.Trust import Annex.UUID +import Annex.VectorClock import qualified Remote import Utility.HumanTime @@ -70,15 +71,15 @@ start (Expire expire) noact actlog descs u = where lastact = changed <$> M.lookup u actlog whenactive = case lastact of - Just (Date t) -> do - d <- liftIO $ durationSince $ posixSecondsToUTCTime t + Just (VectorClock c) -> do + d <- liftIO $ durationSince $ posixSecondsToUTCTime c return $ "last active: " ++ fromDuration d ++ " ago" _ -> return "no activity" desc = fromUUID u ++ " " ++ fromMaybe "" (M.lookup u descs) notexpired ent = case ent of Unknown -> False - Date t -> case lookupexpire of - Just (Just expiretime) -> t >= expiretime + VectorClock c -> case lookupexpire of + Just (Just expiretime) -> c >= expiretime _ -> True lookupexpire = headMaybe $ catMaybes $ map (`M.lookup` expire) [Just u, Nothing] diff --git a/Command/Forget.hs b/Command/Forget.hs index 583eee7ca..d172cc693 100644 --- a/Command/Forget.hs +++ b/Command/Forget.hs @@ -11,8 +11,7 @@ import Command import qualified Annex.Branch as Branch import Logs.Transitions import qualified Annex - -import Data.Time.Clock.POSIX +import Annex.VectorClock cmd :: Command cmd = command "forget" SectionMaintenance @@ -36,10 +35,10 @@ seek = commandAction . start start :: ForgetOptions -> CommandStart start o = do showStart "forget" "git-annex" - now <- liftIO getPOSIXTime - let basets = addTransition now ForgetGitHistory noTransitions + c <- liftIO currentVectorClock + let basets = addTransition c ForgetGitHistory noTransitions let ts = if dropDead o - then addTransition now ForgetDeadRemotes basets + then addTransition c ForgetDeadRemotes basets else basets next $ perform ts =<< Annex.getState Annex.force diff --git a/Command/MetaData.hs b/Command/MetaData.hs index 617b291a1..d10fc9921 100644 --- a/Command/MetaData.hs +++ b/Command/MetaData.hs @@ -9,6 +9,7 @@ module Command.MetaData where import Command import Annex.MetaData +import Annex.VectorClock import Logs.MetaData import Annex.WorkTree import Messages.JSON (JSONActionItem(..)) @@ -18,7 +19,6 @@ import qualified Data.Set as S import qualified Data.Map as M import qualified Data.Text as T import qualified Data.ByteString.Lazy.UTF8 as BU -import Data.Time.Clock.POSIX import Data.Aeson import Control.Concurrent @@ -68,28 +68,28 @@ optParser desc = MetaDataOptions seek :: MetaDataOptions -> CommandSeek seek o = case batchOption o of NoBatch -> do - now <- liftIO getPOSIXTime + c <- liftIO currentVectorClock let seeker = case getSet o of Get _ -> withFilesInGit GetAll -> withFilesInGit Set _ -> withFilesInGitNonRecursive "Not recursively setting metadata. Use --force to do that." withKeyOptions (keyOptions o) False - (startKeys now o) - (seeker $ whenAnnexed $ start now o) + (startKeys c o) + (seeker $ whenAnnexed $ start c o) (forFiles o) Batch -> withMessageState $ \s -> case outputType s of JSONOutput _ -> batchInput parseJSONInput $ commandAction . startBatch _ -> giveup "--batch is currently only supported in --json mode" -start :: POSIXTime -> MetaDataOptions -> FilePath -> Key -> CommandStart -start now o file k = startKeys now o k (mkActionItem afile) +start :: VectorClock -> MetaDataOptions -> FilePath -> Key -> CommandStart +start c o file k = startKeys c o k (mkActionItem afile) where afile = AssociatedFile (Just file) -startKeys :: POSIXTime -> MetaDataOptions -> Key -> ActionItem -> CommandStart -startKeys now o k ai = case getSet o of +startKeys :: VectorClock -> MetaDataOptions -> Key -> ActionItem -> CommandStart +startKeys c o k ai = case getSet o of Get f -> do l <- S.toList . currentMetaDataValues f <$> getCurrentMetaData k liftIO $ forM_ l $ @@ -97,14 +97,14 @@ startKeys now o k ai = case getSet o of stop _ -> do showStart' "metadata" k ai - next $ perform now o k + next $ perform c o k -perform :: POSIXTime -> MetaDataOptions -> Key -> CommandPerform -perform now o k = case getSet o of +perform :: VectorClock -> MetaDataOptions -> Key -> CommandPerform +perform c o k = case getSet o of Set ms -> do oldm <- getCurrentMetaData k let m = combineMetaData $ map (modMeta oldm) ms - addMetaData' k m now + addMetaData' k m c next $ cleanup k _ -> next $ cleanup k @@ -169,7 +169,7 @@ startBatch (i, (MetaData m)) = case i of , keyOptions = Nothing , batchOption = NoBatch } - now <- liftIO getPOSIXTime + t <- liftIO currentVectorClock -- It would be bad if two batch mode changes used exactly -- the same timestamp, since the order of adds and removals -- of the same metadata value would then be indeterminate. @@ -178,7 +178,7 @@ startBatch (i, (MetaData m)) = case i of -- probably less expensive than cleaner methods, -- such as taking from a list of increasing timestamps. liftIO $ threadDelay 1 - next $ perform now o k + next $ perform t o k mkModMeta (f, s) | S.null s = DelMeta f Nothing | otherwise = SetMeta f s -- cgit v1.2.3