There are situations where a container running Discourse may want to
cache the critical DNS services without running the cache_critical_dns
service, for example running migrations prior to running a full bore
application container.
Add a `--once` argument for the cache_critical_dns script that will
only execute the main loop once, and return the status code for the
script to use when exiting. 0 indicates no errors occured during SRV
resolution, and 1 indicates a failure during the SRV lookup.
Nothing is reported to prometheus in run_once mode. Generally this
mode of operation would be a part of a unix pipeline, in which the exit
status is a more meaningful and immediate signal than a prometheus metric.
The reporting has been moved into it's own method that can be called
only when the script is running as a service.
See /t/69597.
Describes the behaviour and configuration of the cache_critical_dns
script, mainly cribbed from commit messages. Tries to make this program
a bit less of an enigma.
The `PG::Connection#ping` method is only reliable for checking if the
given host is accepting connections, and not if the authentication
details are valid.
This extends the healthcheck to confirm that the auth details are
able to both create a connection and execute queries against the
database.
We expect the empty query to return an empty result set, so we can
assert on that. If a failure occurs for any reason, the healthcheck will
return false.
This commit migrates all bookmarks to be polymorphic (using the
bookmarkable_id and bookmarkable_type) columns. It also deletes
all the old code guarded behind the use_polymorphic_bookmarks setting
and changes that setting to true for all sites and by default for
the sake of plugins.
No data is deleted in the migrations, the old post_id and for_topic
columns for bookmarks will be dropped later on.
An SRV RR contains a priority value for each of the SRV targets that
are present, ranging from 0 - 65535. When caching SRV records we may want to
filter out any targets above or below a particular threshold.
This change adds support for specifying a lower and/or upper bound on
target priorities for any SRV RRs. Any targets returned when resolving
the SRV RR whose priority does not fall between the lower and upper
thresholds are ignored.
For example: Let's say we are running two Redis servers, a primary and
cold server as a backup (but not a replica). Both servers would pass health
checks, but clearly the primary should be preferred over the backup
server. In this case, we could configure our SRV RR with the primary
target as priority 1 and backup target as priority 10. The
`DISCOURSE_REDIS_HOST_SRV_LE` could then be set to 1 and the target with
priority 10 would be ignored.
See /t/66045.
`ActiveRecord::Base.connection_config` has been deprecated since Rails
6.1 and was completely removed from Rails 7.
Instead we need to use
`ActiveRecord::Base.connection_db_config.configuration_hash`.
Import scripts were forgotten when we did the Rails 7 upgrade, this
patch fixes them.
A bit of a mixed bag, this addresses several edge areas of bookmarks and makes them compatible with polymorphic bookmarks (hidden behind the `use_polymorphic_bookmarks` site setting). The main ones are:
* ExportUserArchive compatibility
* SyncTopicUserBookmarked job compatibility
* Sending different notifications for the bookmark reminders based on the bookmarkable type
* Import scripts compatibility
* BookmarkReminderNotificationHandler compatibility
This PR also refactors the `register_bookmarkable` API so it accepts a class descended from a `BaseBookmarkable` class instead. This was done because we kept having to add more and more lambdas/properties inline and it was very messy, so a factory pattern is cleaner. The classes can be tested independently as well.
Some later PRs will address some other areas like the discourse narrative bot, advanced search, reports, and the .ics endpoint for bookmarks.
This removes the option to override the sleep time between caching of
DNS records. The override was invalid because `''.to_i` is 0 in Ruby,
causing a tight loop calling the `run` method.
For Redis connections that operate over TLS, we need to ensure that we
are setting the correct arguments for the Redis client. We can utilise
the existing environment variable `DISCOURSE_REDIS_USE_SSL` to toggle
this behaviour.
No SSL verification is performed for two reasons:
- the Discourse application will perform a verification against any FQDN
as specified for the Redis host
- the healthcheck is run against the _resolved_ IP address for the Redis
hostname, and any SSL verification will always fail against a direct
IP address
If no SSL arguments are provided, the IP address is never cached against
the hostname as no healthy address is ever found in the HealthyCache.
Modify the cache_critical_dns script for SRV RR awareness. The new
behaviour is only enabled when one or more of the following environment
variables are present (and only for a host where the `DISCOURSE_*_HOST_SRV`
variable is present):
- `DISCOURSE_DB_HOST_SRV`
- `DISCOURSE_DB_REPLICA_HOST_SRV`
- `DISCOURSE_REDIS_HOST_SRV`
- `DISCOURSE_REDIS_REPLICA_HOST_SRV`
Some minor changes in refactor to original script behaviour:
- add Name and SRVName classes for storing resolved addresses for a hostname
- pass DNS client into main run loop instead of creating inside the loop
- ensure all times are UTC
- add environment override for system hosts file path and time between DNS
checks mainly for testing purposes
The environment variable for `BUNDLE_GEMFILE` is set to enables Ruby to
load gems that are installed and vendored via the project's Gemfile.
This script is usually not run from the project directory as it is
configured as a system service (see
71ba9fb7b5/templates/cache-dns.template.yml (L19))
and therefore cannot load gems like `pg` or `redis` from the default
load paths. Setting this environment variable configures bundler to look
in the correct project directory during it's setup phase.
When a `DISCOURSE_*_HOST_SRV` environment variable is present, the
decision for which target to cache is as follows:
- resolve the SRV targets for the provided hostname
- lookup the addresses for all of the resolved SRV targets via the
A and AAAA RRs for the target's hostname
- perform a protocol-aware healthcheck (PostgreSQL or Redis pings)
- pick the newest target that passes the healthcheck
From there, the resolved address for the SRV target is cached against
the hostname as specified by the original form of the environment
variable.
For example: The hostname specified by the `DISCOURSE_DB_HOST` record
is `database.example.com`, and the `DISCOURSE_DB_HOST_SRV` record is
`database._postgresql._tcp.sd.example.com`. An SRV RR lookup will return
zero or more targets. Each of the targets will be queried for A and AAAA
RRs. For each of the addresses returned, the newest address that passes
a protocol-aware healthcheck will be cached. This address is cached so
that if any newer address for the SRV target appears we can perform a
health check and prefer the newer address if the check passes.
All resolved SRV targets are cached for a minimum of 30 minutes in memory
so that we can prefer newer hosts over older hosts when more than one target
is returned. Any host in the cache that hasn't been seen for more than 30
minutes is purged.
See /t/61485.
If someone types `yes` rather than `YES`, continue anyway.
The chance of typing `yes`, when you actually want to stop, is non-existent. The chance of typing `yes` when you meant `YES` is high, and it's very frustrating when the script quite because you got the case wrong!
This commit promotes all post_deploy migrations which existed in Discourse v2.7.13 (timestamp <= 20210328233843)
This reduces the likelihood of issues relating to migration run order
Also fixes a couple of typos in `script/promote_migrations`
This allows text editors to use correct syntax coloring for the heredoc sections.
Heredoc tag names we use:
languages: SQL, JS, RUBY, LUA, HTML, CSS, SCSS, SH, HBS, XML, YAML/YML, MF, ICS
other: MD, TEXT/TXT, RAW, EMAIL
We validate the *format* of email addresses in many places with a match against
a regex, often with very slightly different syntax.
Adding a separate EmailAddressValidator simplifies the code in a few spots and
feels cleaner.
Deprecated the old location in case someone is using it in a plugin.
No functionality change is in this commit.
Note: the regex used at the moment does not support using address literals, e.g.:
* localpart@[192.168.0.1]
* localpart@[2001:db8::1]
* Optional import of custom user fields from phpBB 3.1+
* Optional import of likes from phpBB3
Requires the phpBB "Thanks for posts" extension
* Fix import of bookmarks from phpBB3
* Update `created_at` of existing user
* Support mapping of phpBB forums to existing Discourse categories
This is in addition to the ability of merging phpBB forums and importing into newly created Discourse categories.
1. bbcode hashes don't always have exactly 8 characters.
2. colors aren't always hex values, it can be a color string ("red", "blue", etc).
3. The closing tag of smileys doesn't always include a `:` character (the start of the regex was already right for this particular issue)
* File.exists? is deprecated and removed in Ruby 3.2 in favor of
File.exist?
* Dir.exists? is deprecated and removed in Ruby 3.2 in favor of
Dir.exist?
The discourse base image already contains a postgres installation, so pulling a separate postgres image is a little wasteful. Using the copy of Postgres in the discourse image saves about 20 seconds on every GitHub actions run.
This commit sets up Postgres with a few performance-improving flags, which we were already using for the `rake docker:test` task (used on our internal CI system).
Some tables in the database have constraints on columns with dates. Because of them, the script for moving timestamps can fail from time to time. This PR makes the script work with such tables.
In general, in PostgreSQL it is not always possible to defer constraint checks to the transaction commit (Primary Keys and Unique Constraints can be deferred, but them should be declared as DEFERRABLE to make it possible. Indices created with CREATE UNIQUE INDEX can't be deferred at all).
Since we can't defer constraint checks, I've made it work using a little hack. For example, if we need to move all timestamps by one day, the script will move timestamps by 1000 years and one day, and then return timestamps back by 1000 years. The script use this hack only for columns that have unique constraints.
This will fix the try-reset build that failed today. Probably this going to happen again with other tables that have constraints on date columns. I'm going to modify the script to make it work without ignoring such tables. After that, the only table we're going to need to ignore will be the 2FA table.
Before I fixed that, don't hesitate to tag me if the try-reset build fail again.
Without checking if t.table_schema = '#{@schema}' the SELECT with JOIN in the script were returning every column twice in case there is a 'backup' scheme with exactly the same tables as in the 'public scheme'
We're going to use this script for updating timestamps on Try, but it can be used with a local database during development as well.
Usage:
Commands:
ruby db_timestamp_updater.rb yesterday <date> move all timestamps by x days so that <date> will be moved to yesterday
ruby db_timestamp_updater.rb 100 move all timestamps forward by 100 days
ruby db_timestamp_updater.rb -100 move all timestamps backward by 100 days
The script moves all timestamps in the database by the same amount of days forward or backward. No need to change the script if we add a new column in the future.
The more simple solution would be just to move timestamps in several tables (topics, posts, and so on). I didn't want to go that way because it could generate additional work in the future. For example, if we add a new column with a timestamp and users can see that timestamp we'd need to add that column to the script. Or, for example, if we move a post's timestamp to the future but forget to move a timestamp of topic timer or user action it can cause weird bugs.
Post-deploy migrations exist to allow for seamless Discourse upgrades. By design, they cause migrations to run out of numerical order. This has the potential to cause some unexpected edge cases. To reduce the likelihood of these edge cases, we will promote historical post_deploy migrations to regular migrations after a full Discourse stable release cycle.
This script is intended to be run at least during every Discourse release cycle.
This means that truly seamless upgrades will not be possible between non-consecutive Discourse versions. (Upgrades will still work, but may cause some server errors for users during the upgrade)
Setting a random value in the interval 1 week ago ... now works better
because this spreads digest scheduling over a week because digests are
sent one week from the date of the last digest.
Over the years we accrued many spelling mistakes in the code base.
This PR attempts to fix spelling mistakes and typos in all areas of the code that are extremely safe to change
- comments
- test descriptions
- other low risk areas
This is a pretty straightforward bulk importer, just tailored to the vBulletin 5 database structure.
Also made a few minor improvements to the base importer -- should be self explanatory in the code.
Wrote up a new script to import from Higher Logic. Nothing too crazy going on here. Two major things about this script:
It requires you to convert a Microsoft SQL file to a format MySQL can read.
Higher Logic stores posts (at least in the case of the import I ran) with the email thread shown in the post body. The script does its best to truncate this out, but the logic may need to be improved on future imports. For the import I ran, it worked just fine as is. 🤷♂️
Made some improvements to the Vanilla MySQL script -- mainly because not all SQL imports require use of the VanillaBodyParser. Still left it as an option to turn on and use if so desired. Also added subcategory support, importing of likes, and solve status.
Includes:
* DEV: Remove external plugin linting (that's covered by CI in their repositories)
* DEV: Move lint stages to a separate workflow (partial de-`if`-ication of workflows)
* DEV: Run CI on `main` branch too
* DEV: Update postgres to 13
* DEV: Update redis to 6.x
Other changes:
* DEV: Remove matrix.os
* DEV: Remove env.BUILD_TYPE
* DEV: Remove env.TARGET
* DEV: Rename `build_types` config option to `build_type`
* DEV: Lowercase `target` and `build_type` names
* DEV: Rename `ci` to `tests`
* DEV: Rename `lint` to `linting`
* DEV: Lower the wizard qunit timeout (30 min -> 10)
* DEV: Ruby version is no longer configurable
* DEV: Run plugin tests only in the `plugins` target
* DEV: Use binstubs where applicable
* DEV: We don't open PRs to `tests-passed`
This is an importer I wrote to restore some users that were
accidentally deleted for being purged as old staged users or old
unactivated users.
It reads from CSV files exported from a discourse sql backup.
When running an import script there are many site settings that are
changed but we reset them back to where they were originally before the
import. However, there are two settings that we don't roll back:
```
purge_unactivated_users_grace_period_days
purge_deleted_uploads_grace_period_days
```
which could have some unintended consequences.
My first question is do we *really* have to change these settings? I'm
not a huge fan of changing someones settings without them really knowing
they were changed.
If we really do have to change these settings here is my proposed PR
where we don't alter the `purge_unactivated_users_grace_period_days` if
it has been disabled.
As I'm writing this another change we could make is that we don't change
either of these site settings if we detect that they aren't set to the
default values.
The drive behind this PR is that there is a discourse instance which
relies on staged users as part of their workflow and this setting was
changed by accident via the import script causing users to be deleted
that shouldn't have been.
You can use `discourse restore --location=local FILENAME` if you want to restore a backup that is stored locally even though the `backup_location` has the value `s3`.
The 'Discourse SSO' protocol is being rebranded to DiscourseConnect. This should help to reduce confusion when 'SSO' is used in the generic sense.
This commit aims to:
- Rename `sso_` site settings. DiscourseConnect specific ones are prefixed `discourse_connect_`. Generic settings are prefixed `auth_`
- Add (server-side-only) backwards compatibility for the old setting names, with deprecation notices
- Copy `site_settings` database records to the new names
- Rename relevant translation keys
- Update relevant translations
This commit does **not** aim to:
- Rename any Ruby classes or methods. This might be done in a future commit
- Change any URLs. This would break existing integrations
- Make any changes to the protocol. This would break existing integrations
- Change any functionality. Further normalization across DiscourseConnect and other auth methods will be done separately
The risks are:
- There is no backwards compatibility for site settings on the client-side. Accessing auth-related site settings in Javascript is fairly rare, and an error on the client side would not be security-critical.
- If a plugin is monkey-patching parts of the auth process, changes to locale keys could cause broken error messages. This should also be unlikely. The old site setting names remain functional, so security-related overrides will remain working.
A follow-up commit will be made with a post-deploy migration to delete the old `site_settings` rows.
* FEATURE: Import attachments
* FEATURE: Add support for importing multiple forums in one
* FEATURE: Add support for category and tag mapping
* FEATURE: Import groups
* FIX: Add spaces around images
* FEATURE: Custom mapping of user rank to trust levels
* FIX: Do not fail import if it cannot import polls
* FIX: Optimize existing records lookup
Co-authored-by: Gerhard Schlager <mail@gerhard-schlager.at>
Co-authored-by: Jarek Radosz <jradosz@gmail.com>
After running the Discourse merge script, it was pretty evident it held up well after all these years ;)
Made a few fixes:
Included an environment variable for DB_PASS as likely the password will need to be changed if running the import in an official Docker container (recommended)
Set a hard order for imported categories, otherwise sometimes they'd be imported in a weird order making things unpredictable for parent/child category imports
Fixed a couple of instances where we added unique indexes (such as on category slugs)
Set up upload regex to handle AWS URLs better
Fixed the script to work with frozen string literals
This commit adds an additional find_user_by_email hook to ManagedAuthenticator so that GitHub login can continue to support secondary email addresses
The github_user_infos table will be dropped in a follow-up commit.
This is the last core authenticator to be migrated to ManagedAuthenticator 🎉
* ensure emails don't have spaces
* import banned users as suspended for 1k yrs
* upgrade users to TL2 if they have comments
* topic: import views, closed and pinned info
* import messages
* encode vanilla usernames for permalinks. Vanilla usernames can contain spaces and special characters.
* parse Vanilla's new rich body format
Adjustments to the base:
1. PG connection doesn't require host - it was broken on import droplet
2. Drop `topic_reply_count` - it was removed here - https://github.com/discourse/discourse/blob/master/db/post_migrate/20200513185052_drop_topic_reply_count.rb
3. Error with `backtrace.join("\n")` -> `e.backtrace.join("\n")`
4. Correctly link the user and avatar to quote block
Adjustments to vanilla:
1. Top-level Vanilla categories are valid categories
2. Posts have `format` column which should be used to decide if the format is HTML or Markdown
3. Remove no UTF8 characters
4. Remove not supported HTML elements like `font` `span` `sub` `u`
- Stream the queries that load the imported_ids
- Use an array instead of a hash for keeping the mapping between imported_ids and new ids
- Ensure we always treat the imported_ids as integers instead of strings
Fixed bugs, added specs, extracted the upload downsizing code to a class, added support for non-S3 setups, changed it so that images aren't downloaded twice.
This code has been tested on production and successfully resized ~180k uploads.
Includes:
* DEV: Extract upload downsizing logic
* DEV: Add support for non-S3 uploads
* DEV: Process only images uploaded by users
* FIX: Incorrect usage of `count` and `exist?` typo
* DEV: Spec S3 image downsizing
* DEV: Avoid downloading images twice
* DEV: Update filesizes earlier in the process
* DEV: Return false on invalid upload
* FIX: Download images that currently above the limit (If the image size limit is decreased, then there was no way to resize those images that now fall outside the allowed size range)
* Update script/downsize_uploads.rb (Co-authored-by: Régis Hanol <regis@hanol.fr>)
Refactors script to follow conventions of other importers and adds some features including like import, processing of post raw text, and, if needed, SSO import.
FEATURE: new rake task to update first_post_created_at column
The not-equal operator (`<>`) in PostgreSQL does not compare values
with NULL. We should instead use `IS DISTINCT FROM` when comparing
values with NULL.
Moves the most important checks into a linter. It gets executed by Lefthook as well as the docker rake task and Github actions. Doing those checks in rspec takes too long and it produces errors when the discourse:test Docker image contains old, invalid locale files.
`Facter.reset` (65d167eac9/lib/facter.rb (L126-L137)) clears `Facter::Options[:external_dir]` which seems to be the 4.x equivalent of `Facter::Util::Config.external_facts_dirs`.
This commit also makes sure that version 4.0 or higher is installed.
Correctly handles more upload formats in posts, updates post custom fields, fixes more edge cases, adds debugging capabilities. (VERBOSE=1 and INTERACTIVE=1 flags)
Includes these commits and some more:
* DEV: Show the fixed image dimensions
* FIX: Support more upload url formats
* DEV: Remove the old upload after updating posts
* FIX: Use the `process_post_#{id}` mutex
* FIX: Avoid rebaking twice
* DEV: Print out the link to the post
* DEV: Process posts chronologically
* DEV: Do a dry-run before saving, pause on any issue
* FIX: Also process deleted posts
* DEV: Make matchers case-insensitive
* DEV: Pause on "detached" uploads, add more debug info
* DEV: Print out time when finished
* DEV: Add support for WORKER_ID/WORKER_COUNT
* DEV: Fix the onebox in cooked text heuristic
* DEV: Don't report already processed posts
* DEV: Beep when done!
* DEV: Ignore issues with deleted posts
* DEV: Ignore issues with deleted topics
* DEV: Multiline SQL
* DEV: Use the bulk attribute assignment
* DEV: Add ENV["INTERACTIVE"] mode
* DEV: Handle post custom fields
* DEV: Bail on non-S3 sites
* DEV: Allow sizes smaller than 1 mpix
* PERF: Dematerialize topic_reply_count
It's only ever used for trust level promotions that run daily, or compared to 0. We don't need to track it on every post creation.
* UX: Add symbol in TL3 report if topic reply count is capped
* DEV: Drop user_stats.topic_reply_count column
* DEV: api documentation updates
- Created a script to convert json responses to rswag
- Documented several api endpoints
- Switched rswag to use header based auth
* Update script, fix some schema missmatches
We have the `# frozen_string_literal: true` comment on all our
files. This means all string literals are frozen. There is no need
to call #freeze on any literals.
For files with `# frozen_string_literal: true`
```
puts %w{a b}[0].frozen?
=> true
puts "hi".frozen?
=> true
puts "a #{1} b".frozen?
=> true
puts ("a " + "b").frozen?
=> false
puts (-("a " + "b")).frozen?
=> true
```
For more details see: https://samsaffron.com/archive/2018/02/16/reducing-string-duplication-in-ruby
* DEV: Update the working tree just once.
`git pull` was effectively doing `git fetch` and `git merge FETCH_HEAD`, and only then we were checking out the desired branch/commit. This change will skip the the merge step.
* DEV: Don't run lefthook in docker_test
I'm not clear why changing only the `wait_for_url` address was necessary and not also the `get` a few lines above, but this change seems to work for me on both literatecomputing.com Groups and a public group.
Checking if all records have been imported uses a temp table in PostgreSQL. This fails when pgbouncer is used unless the temp table is created inside a transaction.
* Detects mostly all attachments and it's a lot faster
* Parses user properties in Ruby instead of the DB, because that's less errorprone
* Imports user avatars
* Imports topic views by users
* Better handling of quotes and YouTube links
* Adds ability to map forums to categories and tags as well as ignore forums.
* Fixes regular expression for detecting attachments in posts.
* Handles "remote attachments" 😮 by inserting a link.
* Imports view counts for topics.
* Handles incorrect references of parent posts.
* Better handling of quotes.
* Finds a lot more attachments by trying to replace various Unicode characters in filenames.
* Customizable email subject prefixes to remove "Re" and "Fwd" as well as localized prefixes.
* Configuration option for prefixes like [FOO] or (BAR) which can be replaced with tags during import.
* Bugfix: Import script might have skipped some users due to missing ORDER BY.
Posts without a user probably shouldn't happen unless there was some direct database tampering, but data like that has been seen in the wild.
The importer will assign those posts to the "system" user.
Previously we had many places in the app that called `hostname` to get
hostname of a server. This commit replaces the pattern in 2 ways
1. We cache the result in `Discourse.os_hostname` so it is only ever called once
2. We prefer to use Socket.gethostname which avoids making a shell command
This improves performance as we are not spawning hostname processes throughout
the app lifetime
This is not used in core or official plugins, and has been printing a deprecation notice since v2.3.0beta4. All OpenID 2.0 code and dependencies have been dropped. The user_open_ids table remains for now, in case anyone has missed the deprecation notice, and needs to migrate their data.
Context at https://meta.discourse.org/t/-/113249
With this change the script:
* Actually removes original large-sized images
* Doesn't save processed files if their size has increased
* Prevents inconsistent state
We pick the first topic with 30 responses as our bench topic.
Previously we simply picked the last topic, but hand no guarantee on ordering.
This also attempts to correct previous runs of the bench.
The following methods have long been deprecated in ruby due to flaws in their implementation per http://blade.nagaokaut.ac.jp/cgi-bin/vframe.rb/ruby/ruby-core/29293?29179-31097:
URI.escape
URI.unescape
URI.encode
URI.unencode
escape/encode are just aliases for one another. This PR uses the Addressable gem to replace these methods with its own encode, unencode, and encode_component methods where appropriate.
I have put all references to Addressable::URI here into the UrlHelper to keep them corralled in one place to make changes to this implementation easier.
Addressable is now also an explicit gem dependency.
