Summary:
A user reported this stack trace:
http://pastebin.com/6auGbZsE
...on this GitHub issue:
https://github.com/facebook/phabricator/issues/389#issuecomment-36612511
The problem is similar to the original report, but not identical. In this case, we're following a sequence of steps like:
- Run setup checks.
- Check for enabled providers, in order to raise "no providers configured yet" warning.
- Try to generate login/redirect URIs.
- Build the request.
- Set the default base URI.
- Run normal code.
Since we try to generate URIs before we provide a default, this fatals. Instead, don't try to build objects.
An alternative fix might be to try to set defaults earlier, but we depend on some config and on building the Request in order to be able to figure out if a request is HTTP or HTTPS right now. We could assume one, or guess, or use protocol-relative URIs (`///host.com`), but I think this fix is a little cleaner overall. If we keep hitting similar stuff, we could look into alternate fixes.
We could also set some kind of "setup mode" flag and make `getURI()` if it's called during setup mode to detect these during testing. I'd like to hit one more of these before doing that, though.
Test Plan: Reproduced the issue, applied the patch, verified this fixes it.
Reviewers: btrahan
Reviewed By: btrahan
CC: aran
Differential Revision: https://secure.phabricator.com/D8395
Summary: Ref T2222. This will probably have some rough edges for a bit (e.g., weird cases I didn't remember or think of), but there's no change to the underlying data and we can easily revert if things get too messy.
Test Plan: Looked at a variety of revisions and saw sensible output.
Reviewers: btrahan
Reviewed By: btrahan
CC: aran
Maniphest Tasks: T2222
Differential Revision: https://secure.phabricator.com/D8361
Summary:
See D8308. Enabling STRICT_ALL_TABLES prevents this entire class of error, by fataling on truncation instead of truncating. We never want truncation; it is always bad and sometimes extremely bad.
We've recommended this mode for developer installs for a long time, and some users run with it enabled, so it's very unlikely to cause any issues (I've had it enabled locally for at least 6-8 months, I think).
Test Plan:
- Disabled mode.
- Saw warning.
- Enabled mode.
- No warning.
{F117040}
Reviewers: btrahan, chad
Reviewed By: chad
CC: chad, aran, arice
Differential Revision: https://secure.phabricator.com/D8309
Summary:
Ref T4420. Tokenizers currently operate in "preload" or "ondemand" modes. In the former mode, which is default, they'll try to load the entire result list when a page loads.
The theory here was that this would slightly improve the experience for small installs, and once they got big enough they could switch to "ondemand". In practice, several issues have arisen:
- We generally don't have a good mechanism for telling installs that they should tweak perf config -- `metamta.send-immediately` is the canonical example here. Some large installs are probably affected negatively by not knowing to change this setting, and having settings like this is generally annoying.
- We have way way too much config now.
- With the advent of ApplicationSearch, pages like Maniphest make many redundant loads to prefill sources like projects. Most of the time, this data is not used. It's far simpler to switch everything to ondemand than try to deal with this, and dealing with this would mean creating two very complex divergent pathways in the codebase for a mostly theoretical performance benefit which only impacts tiny installs.
- We've been using `tokenizer.ondemand` forever on `secure.phabricator.com` since we have many thousands of user accounts, and it doesn't seem sluggish and works properly.
Removing this config is an easy fix which makes the codebase simpler.
I've retained the ability to use preloaded sources, since they may make sense in some cases (in at least one case -- task priorities -- adding a static source pathway might make sense), and they're part of Javelin itself. However, the code will no longer ever go down that pathway.
Test Plan: Used `secure.phabricator.com` for years with this setting enabled.
Reviewers: btrahan, chad
Reviewed By: btrahan
CC: aran
Maniphest Tasks: T4420
Differential Revision: https://secure.phabricator.com/D8232
Summary:
Ref T4310. Fixes T3720. This change:
- Removes concurrent session limits. Instead, unused sessions are GC'd after a while.
- Collapses all existing "web-1", "web-2", etc., sessions into "web" sessions.
- Dramatically simplifies the code for establishing a session (like omg).
Test Plan: Ran migration, checked Sessions panel and database for sanity. Used existing session. Logged out, logged in. Ran Conduit commands.
Reviewers: btrahan
Reviewed By: btrahan
CC: aran
Maniphest Tasks: T4310, T3720
Differential Revision: https://secure.phabricator.com/D7978
Summary:
Ref T3857.
- Always send mail via daemons. This lets us get rid of this config, and is generally much more performant.
- After D7964, we warn if daemons aren't running.
Test Plan: Sent some mail.
Reviewers: btrahan
Reviewed By: btrahan
CC: aran
Maniphest Tasks: T3857
Differential Revision: https://secure.phabricator.com/D7965
Summary:
Currently, we try to mostly-kind-of-work if daemons aren't running (for example, we send mail in-process). I want to stop doing this. A major motivator is that `metamta.send-immediately` is confusing for a lot of users and frequently the cause of performance problems. Increasingly, functionality of applications depends on the daemons (Harbormaster, Drydock, Nuance all require daemons to do anything at all). They're also fairly stable/robust/well-tested and no reasonable install should be running without them.
This will let us simplify or remove some flags (like `metamta.send-immediately`) and simplify some other processes like search indexing.
Test Plan: Stopped daemons, loaded warnings, saw daemon warning. Started daemons, reloade, no warning.
Reviewers: btrahan
Reviewed By: btrahan
CC: aran
Maniphest Tasks: T3857
Differential Revision: https://secure.phabricator.com/D7964
Summary: Ref T4222. This was used by Facebook while developing Releeph, but should no longer be necessary since Releeph is in the upstream. I can't get an answer out of Facebook about whether they still use it or not (see T4227), so nuke it. We're going to replace it with a more general mechanism (see T4222).
Test Plan: Regenerated celerity map. Browsed some pages, still got resources.
Reviewers: btrahan, hach-que
Reviewed By: hach-que
CC: aran
Maniphest Tasks: T4222
Differential Revision: https://secure.phabricator.com/D7863
Summary:
Fixes T3034. This is obsoleted by modern policies.
This was written by a Facebook intern and is rarely used -- the Hive install might be the only use in the wild. It has never really worked correctly.
Test Plan: `grep`; browsed Differential.
Reviewers: btrahan
Reviewed By: btrahan
CC: aran
Maniphest Tasks: T3034
Differential Revision: https://secure.phabricator.com/D7568
Summary: I'm planning to add more detailed info to Diffusion itself, but catch the big issue here.
Test Plan: Hit config issue locally, then resolved it.
Reviewers: btrahan
Reviewed By: btrahan
CC: aran
Differential Revision: https://secure.phabricator.com/D7439
Summary: We've had support for this for a long time, but it was conditional on config. Since it more-or-less actually does something now, just enable it unconditionally.
Test Plan: Settings -> SSH Public Keys
Reviewers: btrahan
Reviewed By: btrahan
CC: hach-que, aran
Maniphest Tasks: T2230
Differential Revision: https://secure.phabricator.com/D7426
Summary:
Gets rid of some old Differential-specific nonsense and replaces it with general runtime-pluggable Remarkup rules.
Facebook: This removes two options which may be in use. Have any classes being added via config here just subclass the new abstract bases instead. This should take 5 seconds to fix. You can adjust order by overriding `getPriority()` on the rules, if necessary.
Test Plan: See comments.
Reviewers: btrahan
Reviewed By: btrahan
CC: FacebookPOC, andrewjcg, aran
Differential Revision: https://secure.phabricator.com/D7393
Summary: Ref T3794. Drop auxiliary field, use standard field.
Test Plan: Performed migration, field seemed to survive it intact. Edited and viewed tasks.
Reviewers: btrahan
Reviewed By: btrahan
CC: aran
Maniphest Tasks: T3794
Differential Revision: https://secure.phabricator.com/D7036
Summary: Ref T418. Maniphest has an obsolete class-based field selector. Replace it with CustomField-based selectors, which use the nice config UI and are generally way easier to use.
Test Plan: Added custom fields; edited and viewed custom fields on tasks. Everything worked as expected.
Reviewers: btrahan
Reviewed By: btrahan
CC: aran
Maniphest Tasks: T418
Differential Revision: https://secure.phabricator.com/D6998
Summary: See IRC. This is dumb but I think we should try to work by default on Debian, and it doesn't cost us too much. See inline comment for more.
Test Plan:
- No `disable_functions`, restarted, worked fine.
- Set `disable_functions = pcntl_derp`, restarted, worked fine.
- Set `disable_functions = derp`, restarted, setup fatal.
Reviewers: btrahan
Reviewed By: btrahan
CC: aran
Differential Revision: https://secure.phabricator.com/D6741
Summary:
Fixes T3709. PHP has two configuration options ('disable_functions', 'disable_classes') which allow functions and classes to be blacklisted at runtime.
Since these break things in an unclear way, raise a setup fatal if they are set.
We take a slightly more tailored approach to these in `phd` already, but I'd rather try just saying "no, this is bad" and see if we can get away with it. I suspect we can, and there's no legitimate reason to blacklist functions given that Phabricator must have access to, e.g., `proc_open()`.
Test Plan: {F54058}
Reviewers: btrahan
Reviewed By: btrahan
CC: aran
Maniphest Tasks: T3709
Differential Revision: https://secure.phabricator.com/D6739
Summary:
See discussion in T2715. Currently, PHIDs are all hard coded in the PHID application. In the long run, we need to move them out into actual applications.
A specific immediate issue is Releeph, which uses a very very old and very broken mechanism to inject PHIDs in a way that only sort of works.
Moving forward, every PHID type will be provided by a `PhabricatorPHIDType` subclass, which will manage loading it, etc.
This also moves toward cleaning up the "load objects by name" (where "name" means something like `D12`) code, which is an //enormous// mess and spread across at least 4-5 callsites.
Test Plan: Used `phid.lookup` and `phid.query` to load Slowvotes.
Reviewers: btrahan
Reviewed By: btrahan
CC: aran
Differential Revision: https://secure.phabricator.com/D6502
Summary:
These versions are broken, but package distros seem to be picking them up. :/
Since the error you get is completely useless, fatal immediately with a useful message.
Ref T2594.
Test Plan: Faked verisions and hit the issue.
Reviewers: btrahan
Reviewed By: btrahan
CC: brennantaylor, Arijit, aran
Maniphest Tasks: T2594
Differential Revision: https://secure.phabricator.com/D6415
Summary: Fixes T3501. `apc.stat` should generally be 0 in production and 1 in development. Raise a setup warning if it isn't.
Test Plan:
Hit both setup warnings.
{F49176}
Reviewers: btrahan
Reviewed By: btrahan
CC: aran
Maniphest Tasks: T3501
Differential Revision: https://secure.phabricator.com/D6376
Summary: Fixes T3400. Users are crafty. Attempt to outwit them.
Test Plan: Added all kinds of nonsense to my PATH to hit all the errors. Verified sensible-looking error messages which I couldn't figure out any way to misread or outwit.
Reviewers: chad, btrahan
Reviewed By: chad
CC: aran
Maniphest Tasks: T3400
Differential Revision: https://secure.phabricator.com/D6318
Summary:
Fixes T3436. Currently, when installs have configuration options which we don't know about, we raise a fairly confusing/ambiguous message about the options being unknown. Instead:
- Keep a list of previously valid (but now deleted) config, with explanatory reasons for what happened to it. Present this information, along with altenate wording ("Obsolete Config" instead of "Unknown Config") where applicable.
- Show a list of all the places the config is defined.
- Provide an active link to delete it from the web UI.
- Provide a command to delete it from the CLI.
- Allow `bin/config delete` to delete configuration options which no longer have a definition.
Test Plan:
- Set an auth key in database, local and file config.
- Walked through the setup issue, cleaning it up.
- Set an invalid key and made sure I still got a reasonable error (this now has better cleanup instructions).
Reviewers: btrahan, chad
Reviewed By: chad
CC: aran
Maniphest Tasks: T3436
Differential Revision: https://secure.phabricator.com/D6317
Summary: Ref T1536. This is the last major migration. Moves us over to the DB and drops all the config stuff.
Test Plan:
- Ran the migration.
- Saw all my old config brought forward and respected, with accurate settings.
- Ran LDAP import.
- Grepped for all removed config options.
Reviewers: btrahan, chad
Reviewed By: btrahan
CC: aran, wez
Maniphest Tasks: T1536
Differential Revision: https://secure.phabricator.com/D6243
Summary:
Ref T1536. Currently, when you install Phabricator you're dumped on the login screen and have to consult the documentation to learn about `bin/accountadmin`.
Instead, detect that an install is running first-time setup:
- It has no configured providers; and
- it has no user accounts.
We can safely deduce that such an install isn't configured yet, and let the user create an admin account from the web UI.
After they login, we raise a setup issue and lead them to configure authentication.
(This could probably use some UI and copy tweaks.)
Test Plan:
{F46738}
{F46739}
Reviewers: chad, btrahan
Reviewed By: chad
CC: aran
Maniphest Tasks: T1536
Differential Revision: https://secure.phabricator.com/D6228
Summary:
Ref T3354. There's no way for us to test most of the config options which actually affect this limit, so the Phabricator config is basically a canary value to indicate "the administrator hasn't configured anything yet".
Raise a setup issue if it isn't set. There's a trail to get here from Files, but we've de-emphasized the old-school upload form so it's hard to unearth.
Emphasize the warning that you need to read the documentation and configure like 30 other things to make this work.
Test Plan: Cleared my config, verified I got the issue, read it, set my config, issue went away.
Reviewers: jamesr, chad
Reviewed By: chad
CC: aran
Maniphest Tasks: T3354
Differential Revision: https://secure.phabricator.com/D6185
Summary:
These are a bit tricky because we don't want to require you to install a VCS you don't use just to use Phabricator. Test that repositories exist before performing the checks.
I'll couple this with additional checks during repository creation.
Test Plan: Changed binary names to nonexistent ones, verified setup issues raised properly.
Reviewers: btrahan
Reviewed By: btrahan
CC: aran
Differential Revision: https://secure.phabricator.com/D6040
Summary:
Spent an hour or two helping a user figure this out. Make sure I never do that again.
If the webserver is configured with an empty or bogus PATH, binaries like 'which' and 'diff' (and 'git', and 'svn', etc.) may not be available. In most cases, this is fine, because we get an error like "sh: whatever-command not found", which is obvious to diagnose.
In the case of 'diff', we don't get this, because 'diff' is expected to exit with a nonzero code for differing files -- so we interpret the "sh: whatever-command not found" as "files differ" and then try to parse the empty output.
Explicitly check for 'which' (on Windows, 'where') and 'diff' during setup (I plan to refine the behavior around 'git', 'svn' and 'hg' at some point, but this is less pressing since the errors are trivial to support).
Test Plan: Faked failures on all modes, verified setup warnings look reasonable.
Reviewers: btrahan, chad
Reviewed By: btrahan
CC: aran
Differential Revision: https://secure.phabricator.com/D6008
Summary:
See <https://github.com/facebook/phabricator/issues/320>. We have a soft dependency on 'fileinfo', which we try to recover from (with `file`) but won't be able to on Windows and apparently FreeBSD systems. Since users can ignore setup checks anyway now, just raise a warning during install.
I believe almost all installs should have this extension, it has been part of the core for a long time.
Test Plan: Faked setup failure, looked at warning. "Solved" setup failure, saw it go away.
Reviewers: btrahan
Reviewed By: btrahan
CC: aran
Differential Revision: https://secure.phabricator.com/D5952
Summary:
Use correct spelling of 'environment.append-paths' so that the current
value of the variable will display as expected in the
'pygmentize Not Found' setup issue screen.
Test Plan:
* Enabled Pygments but haven't installed it
* Follow 'unresolved setup issues' link to 'Not Found' screen
* See that 'envinronment.append-paths' is None
* Set 'environment.append-paths'
* See that 'envinronment.append-paths' is still None
* Apply this fix
* See that 'environment.append-paths' is now '/usr/bin'
Reviewers: epriestley
Reviewed By: epriestley
CC: aran, Korvin
Differential Revision: https://secure.phabricator.com/D5555
Summary: We use mysqli if it's available by default. Don't require installs to build with mysql.
Test Plan: Applied to new secure.phabricator.com install.
Reviewers: chad, vrana
Reviewed By: chad
CC: aran
Differential Revision: https://secure.phabricator.com/D5194
Summary: hehehe
Test Plan: Reloaded /config/, no more bogus setup issuse.
Reviewers: kwadwon, staticshock, chad
Reviewed By: chad
CC: aran
Differential Revision: https://secure.phabricator.com/D5158
Summary: Check if pygmentize is runnable if pygments is enabled
Test Plan: Enable pygments with pygmentize unavailable in path
Reviewers: epriestley
Reviewed By: epriestley
CC: aran, Korvin
Differential Revision: https://secure.phabricator.com/D5157
Summary: T2381
Test Plan:
Include existing setup issues in the ignore config option,
reduces the number of setup issues in the status bar, moves ignored
issues to the bottom of the list, and marks them as ignored.
Also include a string corresponding to no setup issue, and verify that
application does not break.
Reviewers: epriestley
Reviewed By: epriestley
CC: aran, Korvin
Differential Revision: https://secure.phabricator.com/D5072
Summary: Route all `$_SERVER['HTTP_...']` stuff through AphrontRequest (it would be nice to make this non-static, but the stack is a bit tangled right now...)
Test Plan: Verified CSRF and cascading profiling. `var_dump()`'d User-Agent and Referer and verified they are populated and returned correct values when accessed. Restarted server to trigger setup checks.
Reviewers: vrana
Reviewed By: vrana
CC: aran
Differential Revision: https://secure.phabricator.com/D4888
Summary: Preserving animation of GIF profile Pictures
Test Plan: Uploaded Animated images as profile pictures to check if the animation of gif images is preserved and it does :) somewhat !
Reviewers: epriestley
Reviewed By: epriestley
CC: aran, Korvin
Differential Revision: https://secure.phabricator.com/D4833
Summary:
- Make the warning describe rationale and point at the MySQL manual explicitly.
- Add a reference to the developer mode config, in case the user wants to resolve the probelm by disabling developer mode.
- Now that the message is huge, provide a summary.
- Move from "Database" to "MySQL" setup checks -- this is kind of arbitrary, but the former is used for fatals (pre-install) and the latter for warnings (post-install) right now. This has no practical impact on anything and is purely stylistic.
Test Plan:
{F31798}
{F31799}
Reviewers: edward, blc
Reviewed By: edward
CC: aran
Differential Revision: https://secure.phabricator.com/D4835
Summary: Suggest the MySQL mode STRICT_ALL_TABLES during setup if it is not set. Small improvement to the phabricator.developer-mode comments.
Test Plan: Set the global sql_mode to include or exclude STRICT_ALL_TABLES and check for desired behavior.
Reviewers: epriestley
Reviewed By: epriestley
CC: aran, Korvin
Differential Revision: https://secure.phabricator.com/D4803
Summary: These are pretty straightforward, they just have a fair amount of instructional text with inline markup.
Test Plan: Added and viewed a UIExample.
Reviewers: vrana
Reviewed By: vrana
CC: aran
Maniphest Tasks: T2432
Differential Revision: https://secure.phabricator.com/D4686
Summary: Add installation check for a dot in the domain, which is necessary for some browsers to set cookies.
Test Plan: Restart web server to force the setup procedures to run again.
Reviewers: epriestley
Reviewed By: epriestley
CC: aran, Korvin
Differential Revision: https://secure.phabricator.com/D4710
Summary: Port the database checks over.
Test Plan: Triggered all the checks via intentional misconfiguration.
Reviewers: chad, btrahan
Reviewed By: btrahan
CC: aran
Maniphest Tasks: T2228
Differential Revision: https://secure.phabricator.com/D4590
