Summary: Ref T10860. This doesn't change anything, it just separates all this stuff out of `PhabricatorRepository` since I'm planning to add a bit more state to it and it's already pretty big and fairly separable.
Test Plan: Pulled, pushed, browsed Diffusion.
Reviewers: chad
Reviewed By: chad
Maniphest Tasks: T10860
Differential Revision: https://secure.phabricator.com/D15790
Summary:
Ref T4292. We currently synchronize hosted, clustered, Git repositories when we receive an SSH pull or push.
Additionally:
- Synchronize before HTTP reads and writes.
- Synchronize reads before Conduit requests.
We could relax Conduit eventually and allow Diffusion to say "it's OK to give me stale data".
We could also redirect some set of these actions to just go to the up-to-date host instead of connecting to a random host and synchronizing it. However, this potentially won't work as well at scale: if you have a larger number of servers, it sends all of the traffic to the leader immediately following a write. That can cause "thundering herd" issues, and isn't efficient if replicas are in different geographical regions and the write just went to the east coast but most clients are on the west coast. In large-scale cases, it's better to go to the local replica, wait for an update, then serve traffic from it -- particularly given that writes are relatively rare. But we can finesse this later once things are solid.
Test Plan:
- Pushed and pulled a Git repository over HTTP.
- Browsed a Git repository from the web UI.
Reviewers: chad
Reviewed By: chad
Maniphest Tasks: T4292
Differential Revision: https://secure.phabricator.com/D15758
Summary:
Fixes T2451. Several motivations here, from strongest to weakest:
- Currently, getting blame and file content are closely entwined. This makes fixing T9319 more difficult, and I want to fix it. I want to separate blame from content so there's more flexibility in how we approach this issue.
- This makes pursuing T2450 easier, if it turns out to be a meaningful win.
- If we can get a win on blame performance, we can do `arc blame` eventually if we want.
Test Plan:
- Blamed in SVN, Git and Mercurial.
Reviewers: chad
Reviewed By: chad
Maniphest Tasks: T2451
Differential Revision: https://secure.phabricator.com/D14957
Summary:
Ref T4245. Broaden support to include "ABCD", "rABCD", "1234", "R1234", etc.
This doesn't change the old behavior, just accepts more stuff.
Test Plan:
- Browsed Diffusion.
- Made various calls via API console.
Reviewers: chad
Reviewed By: chad
Maniphest Tasks: T4245
Differential Revision: https://secure.phabricator.com/D14928
Summary:
Ref T7803. Ref T5873. I want to drive Conduit through more shared infrastructure, but can't currently add parameters automatically.
Put a `getX()` around the `defineX()` methods so the parent can provide default behaviors.
Also like 60% of methods don't define any special error types; don't require them to implement this method. I want to move away from this in general.
Test Plan:
- Ran `arc unit --everything`.
- Called `conduit.query`.
- Browsed Conduit UI.
Reviewers: btrahan
Reviewed By: btrahan
Subscribers: hach-que, epriestley
Maniphest Tasks: T5873, T7803
Differential Revision: https://secure.phabricator.com/D12380
Summary:
Ref T2783. I think this served two purposes:
- Improving performance in cases where we "know" a repository is local.
- Preventing loops.
It is now obsolete:
- After D11476, refs can almost always resolve on a fast path.
- As T2783 moves forward, we can usually no longer know when a repository is local without actually looking it up -- almost everything is allowed to run anywhere.
- The cluster behavior in D11475 now prevents loops.
Test Plan: `grep`, browsed around. This didn't really do much of anything anymore.
Reviewers: btrahan
Reviewed By: btrahan
Subscribers: epriestley
Maniphest Tasks: T2783
Differential Revision: https://secure.phabricator.com/D11477
Summary:
Fixes T7020. When an external user makes a Conduit request to Diffusion but the repository isn't hosted locally, we need to proxy it.
This also adds a guard layer to prevent requests from getting infinitely proxied inside the cluster.
In "trivial" configurations (where the repository is a service repository, but the service is on the local device) I'm making us always proxy anyway. This basically makes it reasonable to test this stuff (otherwise you'd have to set up two different installs) and this configuration doesn't make much sense in real life (if you're using multiple machines, making one a dedicating daemons+repo box is almost certainly the most reasonable configuration, even for a cluster size of 2).
Test Plan:
- With a service-hosted repository, made Diffusion conduit calls and browsed the UI. Verified requests got proxied once, then resovled.
- With a non-service repository, made Diffusion conduit calls and browsed UI. Verified requests were handled in-process immediately.
Reviewers: btrahan
Reviewed By: btrahan
Subscribers: epriestley
Maniphest Tasks: T7020
Differential Revision: https://secure.phabricator.com/D11475
