We create the reference database in git-clone(1) quite early before
connecting to the remote repository. Given that we do not yet know about
the object format that the remote repository uses at that point in time
the consequence is that the refdb may be initialized with the wrong
object format.
This is not a problem in the context of the files backend as we do not
encode the object format anywhere, and furthermore the only reference
that we write between initializing the refdb and learning about the
object format is the "HEAD" symref. It will become a problem though once
we land the reftable backend, which indeed does require to know about
the proper object format at the time of creation. We thus need to
rearrange the logic in git-clone(1) so that we only initialize the refdb
once we have learned about the actual object format.
As a first step, move listing of remote references to happen earlier,
which also allow us to set up the hash algorithm of the repository
earlier now. While we aim to execute this logic as late as possible
until after most of the setup has happened already, detection of the
object format and thus later the setup of the reference database must
happen before any other logic that may spawn Git commands or otherwise
these Git commands may not recognize the repository as such.
The first Git step where we expect the repository to be fully initalized
is when we fetch bundles via bundle URIs. Funny enough, the comments
there also state that "the_repository must match the cloned repo", which
is indeed not necessarily the case for the hash algorithm right now. So
in practice it is the right thing to detect the remote's object format
before downloading bundle URIs anyway, and not doing so causes clones
with bundle URIs to fail when the local default object format does not
match the remote repository's format.
Unfortunately though, this creates a new issue: downloading bundles may
take a long time, so if we list refs beforehand they might've grown
stale meanwhile. It is not clear how to solve this issue except for a
second reference listing though after we have downloaded the bundles,
which may be an expensive thing to do.
Arguably though, it's preferable to have a staleness issue compared to
being unable to clone a repository altogether.
Signed-off-by: Patrick Steinhardt <ps@pks.im>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
When fetch.bundleURI is set, 'git fetch' downloads bundles from the
given bundle URI before fetching from the specified remote. However,
when using non-file remotes, 'git fetch --all' will launch 'git fetch'
subprocesses which then read fetch.bundleURI and fetch the bundle list
again. We do not expect the bundle list to have new information during
these multiple runs, so avoid these extra calls by un-setting
fetch.bundleURI in the subprocess arguments.
Be careful to skip fetching bundles for the empty bundle string.
Fetching bundles from the empty list presents some interesting test
failures.
Signed-off-by: Derrick Stolee <derrickstolee@github.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
The creationToken heuristic uses a different mechanism for downloading
bundles from the "standard" approach. Specifically: it uses a concrete
order based on the creationToken values and attempts to download as few
bundles as possible. It also modifies local config to store a value for
future fetches to avoid downloading bundles, if possible.
However, if any of the individual bundles has a failed download, then
the logic for the ordering comes into question. It is important to avoid
infinite loops, assigning invalid creation token values in config, but
also to be opportunistic as possible when downloading as many bundles as
seem appropriate.
These tests were used to inform the implementation of
fetch_bundles_by_token() in bundle-uri.c, but are being added
independently here to allow focusing on faulty downloads. There may be
more cases that could be added that result in modifications to
fetch_bundles_by_token() as interesting data shapes reveal themselves in
real scenarios.
Signed-off-by: Derrick Stolee <derrickstolee@github.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
When a bundle list specifies the "creationToken" heuristic, the Git
client downloads the list and then starts downloading bundles in
descending creationToken order. This process stops as soon as all
downloaded bundles can be applied to the repository (because all
required commits are present in the repository or in the downloaded
bundles).
When checking the same bundle list twice, this strategy requires
downloading the bundle with the maximum creationToken again, which is
wasteful. The creationToken heuristic promises that the client will not
have a use for that bundle if its creationToken value is at most the
previous creationToken value.
To prevent these wasteful downloads, create a fetch.bundleCreationToken
config setting that the Git client sets after downloading bundles. This
value allows skipping that maximum bundle download when this config
value is the same value (or larger).
To test that this works correctly, we can insert some "duplicate"
fetches into existing tests and demonstrate that only the bundle list is
downloaded.
The previous logic for downloading bundles by creationToken worked even
if the bundle list was empty, but now we have logic that depends on the
first entry of the list. Terminate early in the (non-sensical) case of
an empty bundle list.
Signed-off-by: Derrick Stolee <derrickstolee@github.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
When a user specifies a URI via 'git clone --bundle-uri', that URI may
be a bundle list that advertises a 'bundle.heuristic' value. In that
case, the Git client stores a 'fetch.bundleURI' config value storing
that URI.
Teach 'git fetch' to check for this config value and download bundles
from that URI before fetching from the Git remote(s). Likely, the bundle
provider has configured a heuristic (such as "creationToken") that will
allow the Git client to download only a portion of the bundles before
continuing the fetch.
Since this URI is completely independent of the remote server, we want
to be sure that we connect to the bundle URI before creating a
connection to the Git remote. We do not want to hold a stateful
connection for too long if we can avoid it.
To test that this works correctly, extend the previous tests that set
'fetch.bundleURI' to do follow-up fetches. The bundle list is updated
incrementally at each phase to demonstrate that the heuristic avoids
downloading older bundles. This includes the middle fetch downloading
the objects in bundle-3.bundle from the Git remote, and therefore not
needing that bundle in the third fetch.
Signed-off-by: Derrick Stolee <derrickstolee@github.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
Bundle providers may organize their bundle lists in a way that is
intended to improve incremental fetches, not just initial clones.
However, they do need to state that they have organized with that in
mind, or else the client will not expect to save time by downloading
bundles after the initial clone. This is done by specifying a
bundle.heuristic value.
There are two types of bundle lists: those at a static URI and those
that are advertised from a Git remote over protocol v2.
The new fetch.bundleURI config value applies for static bundle URIs that
are not advertised over protocol v2. If the user specifies a static URI
via 'git clone --bundle-uri', then Git can set this config as a reminder
for future 'git fetch' operations to check the bundle list before
connecting to the remote(s).
For lists provided over protocol v2, we will want to take a different
approach and create a property of the remote itself by creating a
remote.<id>.* type config key. That is not implemented in this change.
Later changes will update 'git fetch' to consume this option.
Signed-off-by: Derrick Stolee <derrickstolee@github.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
The creationToken heuristic provides an ordering on the bundles
advertised by a bundle list. Teach the Git client to download bundles
differently when this heuristic is advertised.
The bundles in the list are sorted by their advertised creationToken
values, then downloaded in decreasing order. This avoids the previous
strategy of downloading bundles in an arbitrary order and attempting
to apply them (likely failing in the case of required commits) until
discovering the order through attempted unbundling.
During a fresh 'git clone', it may make sense to download the bundles in
increasing order, since that would prevent the need to attempt
unbundling a bundle with required commits that do not exist in our empty
object store. The cost of testing an unbundle is quite low, and instead
the chosen order is optimizing for a future bundle download during a
'git fetch' operation with a non-empty object store.
Since the Git client continues fetching from the Git remote after
downloading and unbundling bundles, the client's object store can be
ahead of the bundle provider's object store. The next time it attempts
to download from the bundle list, it makes most sense to download only
the most-recent bundles until all tips successfully unbundle. The
strategy implemented here provides that short-circuit where the client
downloads a minimal set of bundles.
However, we are not satisfied by the naive approach of downloading
bundles until one successfully unbundles, expecting the earlier bundles
to successfully unbundle now. The example repository in t5558
demonstrates this well:
---------------- bundle-4
4
/ \
----|---|------- bundle-3
| |
| 3
| |
----|---|------- bundle-2
| |
2 |
| |
----|---|------- bundle-1
\ /
1
|
(previous commits)
In this repository, if we already have the objects for bundle-1 and then
try to fetch from this list, the naive approach will fail. bundle-4
requires both bundle-3 and bundle-2, though bundle-3 will successfully
unbundle without bundle-2. Thus, the algorithm needs to keep this in
mind.
A later implementation detail will store the maximum creationToken seen
during such a bundle download, and the client will avoid downloading a
bundle unless its creationToken is strictly greater than that stored
value. For now, if the client seeks to download from an identical
bundle list since its previous download, it will download the
most-recent bundle then stop since its required commits are already in
the object store.
Add tests that exercise this behavior, but we will expand upon these
tests when incremental downloads during 'git fetch' make use of
creationToken values.
Signed-off-by: Derrick Stolee <derrickstolee@github.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
As documented in the bundle URI design doc in 2da14fad8f (docs:
document bundle URI standard, 2022-08-09), the 'creationToken' member of
a bundle URI allows a bundle provider to specify a total order on the
bundles.
Future changes will allow the Git client to understand these members and
modify its behavior around downloading the bundles in that order. In the
meantime, create tests that add creation tokens to the bundle list. For
now, the Git client correctly ignores these unknown keys.
Create a new test helper function, test_remote_https_urls, which filters
GIT_TRACE2_EVENT output to extract a list of URLs passed to
git-remote-https child processes. This can be used to verify the order
of these requests as we implement the creationToken heuristic. For now,
we need to sort the actual output since the current client does not have
a well-defined order that it applies to the bundles.
Signed-off-by: Derrick Stolee <derrickstolee@github.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
When downloading bundles from a git-remote-https subprocess, the bundle
URI logic wants to be opportunistic and download as much as possible and
work with what did succeed. This is particularly important in the "any"
mode, where any single bundle success will work.
If the URI is not available, the git-remote-https process will die()
with a "fatal:" error message, even though that error is not actually
fatal to the super process. Since stderr is passed through, it looks
like a fatal error to the user.
Suppress stderr to avoid these errors from bubbling to the surface. The
bundle URI API adds its own warning() messages on these failures.
Signed-off-by: Derrick Stolee <derrickstolee@github.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
When downloading a list of bundles in "all" mode, Git has no
understanding of the dependencies between the bundles. Git attempts to
unbundle the bundles in some order, but some may not pass the
verify_bundle() step because of missing prerequisites. This is passed as
error messages to the user, even when they eventually succeed in later
attempts after their dependent bundles are unbundled.
Add a new VERIFY_BUNDLE_QUIET flag to verify_bundle() that avoids the
error messages from the missing prerequisite commits. The method still
returns the number of missing prerequisit commits, allowing callers to
unbundle() to notice that the bundle failed to apply.
Use this flag in bundle-uri.c and test that the messages go away for
'git clone --bundle-uri' commands.
Signed-off-by: Derrick Stolee <derrickstolee@github.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
When the content at a given bundle URI is not understood as a bundle
(based on inspecting the initial content), then Git currently gives up
and ignores that content. Independent bundle providers may want to split
up the bundle content into multiple bundles, but still make them
available from a single URI.
Teach Git to attempt parsing the bundle URI content as a Git config file
providing the key=value pairs for a bundle list. Git then looks at the
mode of the list to see if ANY single bundle is sufficient or if ALL
bundles are required. The content at the selected URIs are downloaded
and the content is inspected again, creating a recursive process.
To guard the recursion against malformed or malicious content, limit the
recursion depth to a reasonable four for now. This can be converted to a
configured value in the future if necessary. The value of four is twice
as high as expected to be useful (a bundle list is unlikely to point to
more bundle lists).
To test this scenario, create an interesting bundle topology where three
incremental bundles are built on top of a single full bundle. By using a
merge commit, the two middle bundles are "independent" in that they do
not require each other in order to unbundle themselves. They each only
need the base bundle. The bundle containing the merge commit requires
both of the middle bundles, though. This leads to some interesting
decisions when unbundling, especially when we later implement heuristics
that promote downloading bundles until the prerequisite commits are
satisfied.
Signed-off-by: Derrick Stolee <derrickstolee@github.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
The previous change created the 'git clone --bundle-uri=<uri>' option.
Currently, <uri> must be a filename.
Update copy_uri_to_file() to first inspect the URI for an HTTP(S) prefix
and use git-remote-https as the way to download the data at that URI.
Otherwise, check to see if file:// is present and modify the prefix
accordingly.
Reviewed-by: Josh Steadmon <steadmon@google.com>
Signed-off-by: Derrick Stolee <derrickstolee@github.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
Cloning a remote repository is one of the most expensive operations in
Git. The server can spend a lot of CPU time generating a pack-file for
the client's request. The amount of data can clog the network for a long
time, and the Git protocol is not resumable. For users with poor network
connections or are located far away from the origin server, this can be
especially painful.
Add a new '--bundle-uri' option to 'git clone' to bootstrap a clone from
a bundle. If the user is aware of a bundle server, then they can tell
Git to bootstrap the new repository with these bundles before fetching
the remaining objects from the origin server.
Reviewed-by: Josh Steadmon <steadmon@google.com>
Signed-off-by: Derrick Stolee <derrickstolee@github.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
