This is to avoid inserting a cluster uuid into the location log when
only dead nodes in the cluster contain the content of a key.
One reason why this is necessary is Remote.keyLocations, which excludes
dead repositories from the list. But there are probably many more.
Implementing this was challenging, because Logs.Location importing
Logs.Cluster which imports Logs.Trust which imports Remote.List resulted
in an import cycle through several other modules.
Resorted to making Logs.Location not import Logs.Cluster, and instead
it assumes that Annex.clusters gets populated when necessary before it's
called.
That's done in Annex.Startup, which is run by the git-annex command
(but not other commands) at early startup in initialized repos. Or,
is run after initialization.
Note that is Remote.Git, it is unable to import Annex.Startup, because
Remote.Git importing Logs.Cluster leads the the same import cycle.
So ensureInitialized is not passed annexStartup in there.
Other commands, like git-annex-shell currently don't run annexStartup
either.
So there are cases where Logs.Location will not see clusters. So it won't add
any cluster UUIDs when loading the log. That's ok, the only reason to do
that is to make display of where objects are located include clusters,
and to make commands like git-annex get --from treat keys as being located
in a cluster. git-annex-shell certainly does not do anything like that,
and I'm pretty sure Remote.Git (and callers to Remote.Git.onLocalRepo)
don't either.
And avoid migrate --update/--aply migrating when the new key was already
present in the repository, and got dropped. Luckily, the location log
allows distinguishing from the new key never having been present!
That is mostly useful for --apply because otherwise dropped files would
keep coming back until the old objects were reaped as unused. But it
seemed to make sense to also do it for --update. for consistency in edge
cases if nothing else. One case where --update can use it is when one
branch got migrated earlier, and we dropped the file, and now another
branch has migrated the same file.
Sponsored-by: Jack Hill on Patreon
This can take a lot of memory. I decided to violate the usual rule in
git-annex that it operate in constant memory no matter how many annexed
objects. In this case, it would be hard to be fast without using a big
map of the location logs. The main difficulty here is that there can be
many git-annex branches and it needs to display a consistent view at a
point in time, which means merging information from multiple git-annex
branches.
I have not checked if there are any laziness leaks in this code. It
takes 1 gb to run in my big repo, which is around what I estimated
before writing it.
2 options that are documented are not yet implemented.
Small bug: With eg --when=1h, it will display at 12:00 then 1:10 if the
next change after 12:59 is then. Then it waits until after 2:10 to
display the next change. It ought to wait until after 2:00.
Sponsored-by: Brock Spratlen on Patreon
Factored out overLocationLogs from CmdLine.Seek, which can calculate this
pretty fast even in a large repo. In my big repo, the time to run git-annex
info went up from 1.33s to 8.5s.
Note that the "backend usage" stats are for annexed files in the working
tree only, not all annexed files. This new data source would let that be
changed, but that would be a confusing behavior change. And I cannot
retitle it either, out of fear something uses the current title (eg parsing
the json).
Also note that, while time says "402108maxresident" in my big repo now,
up from "54092maxresident", top shows the RES constant at 64mb, and it
was 48mb before. So I don't think there is a memory leak. I tried using
deepseq to force full evaluation of addKeyCopies and memory use didn't
change, which also says no memory leak. And indeed, not even calling
addKeyCopies resulted in the same memory use. Probably the increased memory
usage is buffering the stream of data from git in overLocationLogs.
Sponsored-by: Brett Eisenberg on Patreon
It would be difficult to make Annex.Branch.files query the unmerged
git-annex branches. Might be possible, similar to what was discussed in
7f6b2ca49c but again I decided to make it
not do anything in that situation to start with before adding such a
complicated thing.
git-annex info uses it when getting info about a repostory. The choices
were to make that fail with an error, or display the info it can, and
change the output slightly for the bits of info it cannot access. While
that is a behavior change, and I want to avoid any behavior changes due
to unmerged git-annex branches in a read-only repo, displaying a message
that is not a number seems unlikely to break anything that was consuming
a number, any worse than throwing an exception would. Probably.
Also git-annex unused --from origin is made to throw an error, but
it would fail later anyway when trying to write to the unused log files.
Sponsored-by: Dartmouth College's Datalad project
* Deal with clock skew, both forwards and backwards, when logging
information to the git-annex branch.
* GIT_ANNEX_VECTOR_CLOCK can now be set to a fixed value (eg 1)
rather than needing to be advanced each time a new change is made.
* Misuse of GIT_ANNEX_VECTOR_CLOCK will no longer confuse git-annex.
When changing a file in the git-annex branch, the vector clock to use is now
determined by first looking at the current time (or GIT_ANNEX_VECTOR_CLOCK
when set), and comparing it to the newest vector clock already in use in
that file. If a newer time stamp was already in use, advance it forward by
a second instead.
When the clock is set to a time in the past, this avoids logging with
an old timestamp, which would risk that log line later being ignored in favor
of "newer" line that is really not newer.
When a log entry has been made with a clock that was set far ahead in the
future, this avoids newer information being logged with an older timestamp
and so being ignored in favor of that future-timestamped information.
Once all clocks get fixed, this will result in the vector clocks being
incremented, until finally enough time has passed that time gets back ahead
of the vector clock value, and then it will return to usual operation.
(This latter situation is not ideal, but it seems the best that can be done.
The issue with it is, since all writers will be incrementing the last
vector clock they saw, there's no way to tell when one writer made a write
significantly later in time than another, so the earlier write might
arbitrarily be picked when merging. This problem is why git-annex uses
timestamps in the first place, rather than pure vector clocks.)
Advancing forward by 1 second is somewhat arbitrary. setDead
advances a timestamp by just 1 picosecond, and the vector clock could
too. But then it would interfere with setDead, which wants to be
overrulled by any change. So it could use 2 picoseconds or something,
but that seems weird. It could just as well advance it forward by a
minute or whatever, but then it would be harder for real time to catch
up with the vector clock when forward clock slew had happened.
A complication is that many log files contain several different peices of
information, and it may be best to only use vector clocks for the same peice
of information. For example, a key's location log file contains
InfoPresent/InfoMissing for each UUID, and it only looks at the vector
clocks for the UUID that is being changed, and not other UUIDs.
Although exactly where the dividing line is can be hard to determine.
Consider metadata logs, where a field "tag" can have multiple values set
at different times. Should it advance forward past the last tag?
Probably. What about when a different field is set, should it look at
the clocks of other fields? Perhaps not, but currently it does, and
this does not seems like it will cause any problems.
Another one I'm not entirely sure about is the export log, which is
keyed by (fromuuid, touuid). So if multiple repos are exporting to the
same remote, different vector clocks can be used for that remote.
It looks like that's probably ok, because it does not try to determine
what order things occurred when there was an export conflict.
Sponsored-by: Jochen Bartl on Patreon
This adds a separate journal, which does not currently get committed to
an index, but is planned to be committed to .git/annex/index-private.
Changes that are regarding a UUID that is private will get written to
this journal, and so will not be published into the git-annex branch.
All log writing should have been made to indicate the UUID it's
regarding, though I've not verified this yet.
Currently, no UUIDs are treated as private yet, a way to configure that
is needed.
The implementation is careful to not add any additional IO work when
privateUUIDsKnown is False. It will skip looking at the private journal
at all. So this should be free, or nearly so, unless the feature is
used. When it is used, all branch reads will be about twice as expensive.
It is very lucky -- or very prudent design -- that Annex.Branch.change
and maybeChange are the only ways to change a file on the branch,
and Annex.Branch.set is only internal use. That let Annex.Branch.get
always yield any private information that has been recorded, without
the risk that Annex.Branch.set might be called, with a non-private UUID,
and end up leaking the private information into the git-annex branch.
And, this relies on the way git-annex union merges the git-annex branch.
When reading a file, there can be a public and a private version, and
they are just concacenated together. That will be handled the same as if
there were two diverged git-annex branches that got union merged.
All callers adjusted to update it themselves.
In Command.ReKey, and Command.SetKey, the cleanup action already did,
so it was updating the log twice before.
This fixes a bug when annex.stalldetection is set, as now
Command.Transferrer can skip updating the location log, and let it be
updated by the calling process.
Eliminate a zombie that was only cleaned up by the later zombie cleanup
code.
This is still not ideal, it would be cleaner if it used conduit or
something, and if the thread gets killed before waiting, it won't stop
the process.
Only remaining zombies are in CmdLine.Seek
Fix support for repositories tuned with annex.tune.branchhash1=true,
including --all not working and git-annex log not displaying anything for
annexed files.
Finally builds (oh the agoncy of making it build), but still very
unmergable, only Command.Find is included and lots of stuff is badly
hacked to make it compile.
Benchmarking vs master, this git-annex find is significantly faster!
Specifically:
num files old new speedup
48500 4.77 3.73 28%
12500 1.36 1.02 66%
20 0.075 0.074 0% (so startup time is unchanged)
That's without really finishing the optimization. Things still to do:
* Eliminate all the fromRawFilePath, toRawFilePath, encodeBS,
decodeBS conversions.
* Use versions of IO actions like getFileStatus that take a RawFilePath.
* Eliminate some Data.ByteString.Lazy.toStrict, which is a slow copy.
* Use ByteString for parsing git config to speed up startup.
It's likely several of those will speed up git-annex find further.
And other commands will certianly benefit even more.
This does not change the overall license of the git-annex program, which
was already AGPL due to a number of sources files being AGPL already.
Legally speaking, I'm adding a new license under which these files are
now available; I already released their current contents under the GPL
license. Now they're dual licensed GPL and AGPL. However, I intend
for all my future changes to these files to only be released under the
AGPL license, and I won't be tracking the dual licensing status, so I'm
simply changing the license statement to say it's AGPL.
(In some cases, others wrote parts of the code of a file and released it
under the GPL; but in all cases I have contributed a significant portion
of the code in each file and it's that code that is getting the AGPL
license; the GPL license of other contributors allows combining with
AGPL code.)
Most of the individual logs are not converted yet, only presense logs
have an efficient ByteString Builder implemented so far. The rest
convert to and from String.
This should make == comparison of UUIDs somewhat faster, and perhaps a
few other operations around maps of UUIDs etc.
FromUUID/ToUUID are used to convert String, which is still used for all
IO of UUIDs. Eventually the hope is those instances can be removed,
and all git-annex branch log files etc use ByteString throughout, for a
real speed improvement.
Note the use of fromRawFilePath / toRawFilePath -- while a UUID usually
contains only alphanumerics and so could be treated as ascii, it's
conceivable that some git-annex repository has been initialized using
a UUID that is not only not a canonical UUID, but contains high unicode
or invalid unicode. Using the filesystem encoding avoids any problems
with such a thing. However, a NUL in a UUID seems extremely unlikely,
so I didn't use encodeBS / decodeBS to avoid their extra overhead in
handling NULs.
The Read/Show instance for UUID luckily serializes the same way for
ByteString as it did for String.
This leaves git annex unused --from remote still using loggedKeysFor
and buffering more than ought to be necessary, but I can't see a way to
improve that.
In Annex.Branch.branch, the (++) was killing laziness.
Rewrote so it streams lazily.
filterM also kills laziness, so made loggedKeys use a Unchecked type,
and check if the key is dead in the seek loop.
Note that loggedKeysFor still buffers, so git-annex info <remote> and
git-annex unused --from remote still use more memory than necessary.
Also removed some unused functions from Annex.Journal.
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.
fsck already special-cased dead keys to make --all not report errors with
them, and it makes sense to also expand that to whereis. I think it makes
sense for dead keys to be skipped by all uses of --all, so mistakes can be
completely forgotten about and not come back to haunt us.
The speed impact of testing if the key is dead is negligible for fsck and
whereis, since they use the location log anyway and it gets cached.
This does slow down a few commands that support --all, in particular
metadata --all runs around 2x as slow. I don't think metadata
--all is often used though. It might slow down copy/move/mirror
--all and get --all.
log --all is not affected (does not use the normal --all machinery).
Dead keys will still be processed by --incomplete, --branch,
--failed, and --key. Although it would be unlikely for a dead key to
ave in incomplete or failed transfer. It seems to make perfect sense for
--branch to process keys on the branch, even if dead.
(fsck's special-casing of dead keys was left in, so if one of these options
causes a dead key to be fscked, there will be a nice message.)
This commit was supported by the NSF-funded DataLad project.
Implemented with no additional overhead of compares etc.
This is safe to do for presence logs because of their locality of change;
a given repo's presence logs are only ever changed in that repo, or in a
repo that has just been actively changing the content of that repo.
So, we don't need to worry about a split-brain situation where there'd
be disagreement about the location of a key in a repo. And so, it's ok to
not update the timestamp when that's the only change that would be made
due to logging presence info.
I hope this doesn't impact speed much -- it does have to pull out a value
from Annex state every time it accesses the branch now.
The test case I dropped has never caught any problems that I can remember,
and would have been rather difficult to convert.
Having one module that knows about all the filenames used on the branch
allows working back from an arbitrary filename to enough information about
it to implement dropping dead remotes and doing other log file compacting
as part of a forget transition.
Avoid ever using read to parse a non-haskell formatted input string.
show :: Key is arguably still show abuse, but displaying Keys as filenames
is just too useful to give up.
Checks location log information, and file contents.
Does not check that numcopies is satisfied, as .gitattributes information
about numcopies is not available in a bare repository. In practice, that
should not be a problem, since fsck is also run in a checkout and will
check numcopies there.
