Mercurial: Changelog

changegroup: port to emitrevisions() (issue5976) We now have a unified API for emitting revision data from a storage backend. It handles sorting nodes and the complicated delta versus revision decisions for us. This commit ports changegroup to that API. There should be no behavior changes for changegroups not using ellipsis. And lack of test changes seems to confirm that. There are some changes for ellipsis mode, however. Before, when sending an ellipsis revision, we would always send a fulltext revision (as opposed to a delta). There was a TODO tracking this open item. One of the things the emitrevisions() API does for us is figure out whether we can safely emit a delta. So, it is now possible for ellipsis revisions to be sent as deltas! (It does this by not assuming parent/ancestor revisions are available and tracking which revisions have been sent out.) Because we eliminated the list of revision delta request objects, performance has improved substantially: $ hg perfchangegroupchangelog before: ! wall 24.348077 comb 24.330000 user 24.140000 sys 0.190000 (best of 3) after: ! wall 18.245911 comb 18.240000 user 18.100000 sys 0.140000 (best of 3) That's a lot of overhead for creating a few hundred thousand Python objects! This is still a little slower than 4.7. Probably due to 23d582ca introducing a type for the revision/delta results. There is potentially room to optimize. But at some point we need to abstract storage in order to support alternate storage backends. Unfortunately that means using a Python data structure to represent results. And unfortunately there is overhead with every new Python object created. Differential Revision: https://phab.mercurial-scm.org/D4725

wireprotov2server: port to emitrevisions() We now have a proper storage API to request data on multiple revisions. We can drop it into wire protocol version 2 with minimal effort. The new API handles pretty much everything we were doing manually to build up the delta request. So we were able to delete a lot of code. As a bonus, wireprotov2 code is no longer accessing some low-level storage APIs. This includes the assumption that a node has an associated numeric revision number! This should make it drastically simpler to implement a server that doesn't have the concept of revision numbers. Differential Revision: https://phab.mercurial-scm.org/D4724

tests: use more complex file storage test The previous test was attempting to to test delta storage behavior. It didn't do a very good job at it because there was a good chance a delta wasn't being used in storage. Let's switch the test to yield a delta in storage so an upcoming change to delegate delta logic to storage has the desired effect. Differential Revision: https://phab.mercurial-scm.org/D4723

revlog: new API to emit revision data I recently refactored changegroup generation code to make it more storage agnostic. I made significant progress. But there is still a bit of work to be done. Specifically: * Changegroup code is looking at low-level storage attributes to influence sorting. Sorting should be done at the storage layer. * The linknode lookup and sorting code for ellipsis is very complicated. * Linknodes are just generally wonky because e.g. file storage doesn't know how to translate a linkrev to a changelog node. * We regressed performance when introducing the request-response objects. Having thought about this problem a bit, I think I've come up with a better interface for emitting revision deltas. This commit defines and implements that interface. See the docstring in repository.py for more info. This API adds 3 notable features over the previous one. First, it defers node ordering to the storage implementation in the common case but allows overriding as necessary. We have a facility for requesting an exact ordering (used in ellipsis mode). We have another facility for storage order (used for changelog). Second, we have an argument specifying assumptions about parents revisions. This can be used to force a fulltext revision when we don't know the receiver has a parent revision to delta against. Third, we can control whether revision data is emitted. This makes the API suitable as a generic "index data retrieval" API as well as for producing revision deltas - possibly in the same operation! The new API is much simpler: we no longer need a complicated "request" object to encapsulate the delta generation request. I'm optimistic this will restore performance loss associated with emitrevisiondeltas(). Storage unit tests for the new API have been implemented. Future commits will port existing consumers of emitrevisiondeltas() to the new API then remove emitrevisiondeltas(). Differential Revision: https://phab.mercurial-scm.org/D4722

changegroup: remove reordering control (BC) This logic - including the experimental bundle.reorder option - was originally added in a8e3931e3fb5 in 2011 and then later ported to changegroup.py. The intent of this option and associated logic is to control the ordering of revisions in deltagroups in changegroups. At the time it was implemented, only changegroup version 1 existed and generaldelta revlogs were just coming into the world. Changegroup version 1 requires that deltas be made against the last revision sent over the wire. Used with generaldelta, this created an impedance mismatch of sorts and resulted in changegroup producers spending a lot of time recomputing deltas. Revision reordering was introduced so outgoing revisions would be sent in "generaldelta order" and producers would be able to reuse internal deltas from storage. Later on, we introduced changegroup version 2. It supported denoting which revision a delta was against. So we no longer needed to sort outgoing revisions to ensure optimal delta generation from the producer. So, subsequent changegroup versions disabled reordering. We also later made the changelog not store deltas by default. And we also made the changelog send out deltas in storage order. Why we do this for changelog, I'm not sure. Maybe we want to preserve revision order across clones? It doesn't really matter for this commit. Fast forward to 2018. We want to abstract storage backends. And having changegroup code require knowledge about how deltas are stored internally interferes with that goal. This commit removes reordering control from changegroup generation. After this commit, the reordering behavior is: * The changelog is always sent out in storage order (no behavior change). * Non-changelog generaldelta revlogs are reordered to always be in DAG topological order (previously, generaldelta revlogs would be emitted in storage order for version 2 and 3 changegroups). * Non-changelog non-generaldelta revlogs are sent in storage order (no behavior change). * There exists no config option to override behavior. The big difference here is that generaldelta revlogs now *always* have their revisions sorted in DAG order before going out over the wire. This behavior was previously only done for changegroup version 1. Version 2 and version 3 changegroups disabled reordering because the interchange format supported encoding arbitrary delta parents, so reordering wasn't strictly necessary. I can think of a few significant implications for this change. Because changegroup receivers will now see non-changelog revisions in DAG order instead of storage order, the internal storage order of manifests and files may differ substantially between producer and consumer. I don't think this matters that much, since the storage order of manifests and files is largely hidden from users. Only the storage order of changelog matters (because `hg log` shows the changelog in storage order). I don't think there should be any controversy here. The reordering of revisions has implications for changegroup producers. Previously, generaldelta revlogs would be emitted in storage order. And in the common case, the internally-stored delta could effectively be copied from disk into the deltagroup delta. This meant that emitting delta groups for generaldelta revlogs would be mostly linear read I/O. This is desirable for performance. With us now reordering generaldelta revlog revisions in DAG order, the read operations may use more random I/O instead of sequential I/O. This could result in performance loss. But with the prevalence of SSDs and fast random I/O, I'm not too worried. (Note: the optimal emission order for revlogs is actually delta encoding order. But the changegroup code wasn't doing that before or after this change. We could potentially implement that in a later commit.) Changegroups in DAG order will have implications for receivers. Previously, receiving storage order might mean seeing a number of interleaved branches. This would mean long delta chains, sparse I/O, and possibly more fulltext revisions instead of deltas, blowing up storage storage. (This is the same set of problems that sparse revlogs aims to address.) With the producer now sending revisions in DAG order, the receiver also stores revisions in DAG order. That means revisions for the same DAG branch are all grouped together. And this should yield better storage outcomes. In other words, sending the reordered changegroup allows the receiver to have better storage order and for the producer to not propagate its (possibly sub-optimal) internal storage order. On the mozilla-unified repository, this change influences bundle generation: $ hg bundle -t none-v2 -a before: time: real 355.680 secs (user 256.790+0.000 sys 16.820+0.000) after: time: real 382.950 secs (user 281.700+0.000 sys 17.690+0.000) before: 7,150,228,967 bytes (uncompressed) after: 7,041,556,273 bytes (uncompressed) before: 1,669,063,234 bytes (zstd l=3) after: 1,628,598,830 bytes (zstd l=3) $ hg unbundle before: time: real 511.910 secs (user 466.750+0.000 sys 32.680+0.000) after: time: real 487.790 secs (user 443.940+0.000 sys 30.840+0.000) 00manifest.d size: source: 274,924,292 bytes before: 304,741,626 bytes after: 245,252,087 bytes .hg/store total file size: source: 2,649,133,490 before: 2,680,888,130 after: 2,627,875,673 We see the bundle size drop. That's probably because if a revlog internally isn't storing a delta, it will choose to delta against the last emitted revision. And on repos with interleaved branches (like mozilla-unified), the previous revision could be an unrelated branch and therefore be a large delta. But with this patch, the previous revision is likely p1 or p2 and a delta should be small. We also see the manifest size drop by ~50 MB. It's worth noting that the manifest actually *increased* in size by ~25 MB in the old strategy and decreased ~25 MB from its source in the new strategy. Again, my explanation for this is that the DAG ordering in the changegroup is resulting in better grouping of revisions in the receiver, which results in more compact delta chains and higher storage efficiency. Unbundle time also dropped. I suspect this is due to the revlog having to work less to compute deltas since the incoming deltas are more optimal. i.e. the receiver spends less time resolving fulltext revisions as incoming deltas bounce around between DAG branches and delta chains. We also see bundle generation time increase. This is not desirable. However, the regression is only significant on the original repository: if we generate a bundle from the repository created from the new, always reordered bundles, we're close to baseline (if not at it with expected noise): $ hg bundle -t none-v2 -a before (original): time: real 355.680 secs (user 256.790+0.000 sys 16.820+0.000) after (original): time: real 382.950 secs (user 281.700+0.000 sys 17.690+0.000) after (new repo): time: real 362.280 secs (user 260.300+0.000 sys 17.700+0.000) This regression is a bit worrying because it will impact serving canonical repositories (that don't have optimal internal storage unless they are reordered - possibly as part of running `hg debugupgraderepo`). However, this regression will only be noticed by very large changegroups. And I'm guessing/hoping that any repository that large is using clonebundles to mitigate server load. Again, sending DAG order isn't the optimal send order for servers: sending in storage-delta order is. But in order to enable storage-optimal send order, we'll need a storage API that handles sorting. Future commits will introduce such an API. Differential Revision: https://phab.mercurial-scm.org/D4721