This is done in as much of a drop-in fashion as possible. Note that
libvips does not support animated PNGs[^1], and as such this
conversion removes support for them as emoji; however, libvips
includes support for webp images, which future commits will take
advantage of.
This removes the MAX_EMOJI_GIF_SIZE limit, since that existed to work
around bugs in Pillow. MAX_EMOJI_GIF_FILE_SIZE_BYTES is fixed to
actually be 128KiB (not 128MiB, as it actually was), and is counted
_after_ resizing, since the point is to limit the amount of data
transfer to clients.
[^1]: https://github.com/libvips/libvips/discussions/2000
Fixes#29632.
The issue description explains this well:
We currently recalculate `currently_used_upload_space_bytes` every file
upload, by dint of calling `flush_used_upload_space_cache` on
save/delete, and then immediately calling
`user_profile.realm.currently_used_upload_space_bytes()` in
`notify_attachment_update`. Since this walks the Attachments table,
recalculating this can take seconds in large realms.
Switch this to using a CountStat, so we don't need to walk significant
chunks of the Attachment table when we upload an attachment. This will
also give us a historical daily graph of usage.
Most importantly, fixes a bug where a realm with a custom
.upload_quota_gb value (set by changing it in the database via e.g.
manage.py shell) would end up having it lowered while upgrading their
plan via the do_change_realm_plan_type function, which used to just set
it to the value implied by the new plan without caring about whether
that isn't lower than the original limit.
The new approach is cleaner since we don't do db queries by
upload_quota_gb so it's nicer to just generate these dynamically, making
changes to our limit-per-plan rules much easier - skipping the need for
migrations.
If the request's `Accept:` header signals a preference for serving
images over text, return an image representing the 404/403 instead of
serving a `text/html` response.
Fixes: #23739.
Previously, this code:
```python3
old_archived_attachments = ArchivedAttachment.objects.annotate(
has_other_messages=Exists(
Attachment.objects.filter(id=OuterRef("id"))
.exclude(messages=None)
.exclude(scheduled_messages=None)
)
).filter(messages=None, create_time__lt=delta_weeks_ago, has_other_messages=False)
```
...protected from removal any ArchivedAttachment objects where there
was an Attachment which had _both_ a message _and_ a scheduled
message, instead of _either_ a message _or_ a scheduled message.
Since files are removed from disk when the ArchivedAttachment rows are
deleted, this meant that if an upload was referenced in two messages,
and one was deleted, the file was permanently deleted when the
ArchivedMessage and ArchivedAttachment were cleaned up, despite being
still referenced in live Messages and Attachments.
Switch from `.exclude(messages=None).exclude(scheduled_messages=None)`
to `.exclude(messages=None, scheduled_messages=None)` which "OR"s
those conditions appropriately.
Pull the relevant test into its own file, and expand it significantly
to cover this, and other, corner cases.
It replaces the "File not found." text with:
"This file does not exist or has been deleted."
At present when a file is deleted it results in a confusing
experience when looking at the "File not found." message.
In order to clarify the situation is not a bug, the message
has been replaced with a better alternative.
Fixes part of Issue #23739.
Subsequent commits will add "on_delete=models.RESTRICT"
relationships, which will result in the Attachment
objects being deleted after Realm has been deleted from
the database.
In order to handle this, we update
get_realm_used_upload_space_cache_key function to accept
realm_id as parameter instead of realm object, so that
the code for flushing the cache works even after the
realm is deleted. This change is fine because eventually
only realm_id is used by this function and there is no
need of the complete realm object.
Failing to remove all of the rules which were added causes action at a
distance with other tests. The two methods were also only used by
test code, making their existence in zerver.lib.rate_limiter clearly
misplaced.
This fixes one instance of a mis-balanced add/remove, which caused
tests to start failing if run non-parallel and one more anonymous
request was added within a rate-limit-enabled block.
Part of splitting creating and editing scheduled messages.
Should be merged with final commit in series. Breaks tests.
Removes `scheduled_message_id` parameter from the create scheduled
message path.
Fixes#25414.
We add Attachment.scheduled_messages relation to track ScheduledMessages
which reference the attachment.
The import bits can be done after merging this, by updating #25345.
Previously, tests that exercised code paths that added local
uploads did not always clean up `settings.LOCAL_UPLOADS_DIR`
after the test was complete.
Updates the `ZulipTestCase` class to remove any local uploads
in the unique `settings.LOCAL_UPLOADS_DIR` in `tearDown` for
all tests.
The Django convention is for __repr__ to include the type and __str__
to omit it. In fact its default __repr__ implementation for models
automatically adds a type prefix to __str__, which has resulted in the
type being duplicated:
>>> UserProfile.objects.first()
<UserProfile: <UserProfile: emailgateway@zulip.com <Realm: zulipinternal 1>>>
Signed-off-by: Anders Kaseorg <anders@zulip.com>
Black 23 enforces some slightly more specific rules about empty line
counts and redundant parenthesis removal, but the result is still
compatible with Black 22.
(This does not actually upgrade our Python environment to Black 23
yet.)
Signed-off-by: Anders Kaseorg <anders@zulip.com>
When file uploads are stored in S3, this means that Zulip serves as a
302 to S3. Because browsers do not cache redirects, this means that
no image contents can be cached -- and upon every page load or reload,
every recently-posted image must be re-fetched. This incurs extra
load on the Zulip server, as well as potentially excessive bandwidth
usage from S3, and on the client's connection.
Switch to fetching the content from S3 in nginx, and serving the
content from nginx. These have `Cache-control: private, immutable`
headers set on the response, allowing browsers to cache them locally.
Because nginx fetching from S3 can be slow, and requests for uploads
will generally be bunched around when a message containing them are
first posted, we instruct nginx to cache the contents locally. This
is safe because uploaded file contents are immutable; access control
is still mediated by Django. The nginx cache key is the URL without
query parameters, as those parameters include a time-limited signed
authentication parameter which lets nginx fetch the non-public file.
This adds a number of nginx-level configuration parameters to control
the caching which nginx performs, including the amount of in-memory
index for he cache, the maximum storage of the cache on disk, and how
long data is retained in the cache. The currently-chosen figures are
reasonable for small to medium deployments.
The most notable effect of this change is in allowing browsers to
cache uploaded image content; however, while there will be many fewer
requests, it also has an improvement on request latency. The
following tests were done with a non-AWS client in SFO, a server and
S3 storage in us-east-1, and with 100 requests after 10 requests of
warm-up (to fill the nginx cache). The mean and standard deviation
are shown.
| | Redirect to S3 | Caching proxy, hot | Caching proxy, cold |
| ----------------- | ------------------- | ------------------- | ------------------- |
| Time in Django | 263.0 ms ± 28.3 ms | 258.0 ms ± 12.3 ms | 258.0 ms ± 12.3 ms |
| Small file (842b) | 586.1 ms ± 21.1 ms | 266.1 ms ± 67.4 ms | 288.6 ms ± 17.7 ms |
| Large file (660k) | 959.6 ms ± 137.9 ms | 609.5 ms ± 13.0 ms | 648.1 ms ± 43.2 ms |
The hot-cache performance is faster for both large and small files,
since it saves the client the time having to make a second request to
a separate host. This performance improvement remains at least 100ms
even if the client is on the same coast as the server.
Cold nginx caches are only slightly slower than hot caches, because
VPC access to S3 endpoints is extremely fast (assuming it is in the
same region as the host), and nginx can pool connections to S3 and
reuse them.
However, all of the 648ms taken to serve a cold-cache large file is
occupied in nginx, as opposed to the only 263ms which was spent in
nginx when using redirects to S3. This means that to overall spend
less time responding to uploaded-file requests in nginx, clients will
need to find files in their local cache, and skip making an
uploaded-file request, at least 60% of the time. Modeling shows a
reduction in the number of client requests by about 70% - 80%.
The `Content-Disposition` header logic can now also be entirely shared
with the local-file codepath, as can the `url_only` path used by
mobile clients. While we could provide the direct-to-S3 temporary
signed URL to mobile clients, we choose to provide the
served-from-Zulip signed URL, to better control caching headers on it,
and greater consistency. In doing so, we adjust the salt used for the
URL; since these URLs are only valid for 60s, the effect of this salt
change is minimal.
Moving `/user_avatars/` to being served partially through Django
removes the need for the `no_serve_uploads` nginx reconfiguring when
switching between S3 and local backends. This is important because a
subsequent commit will move S3 attachments to being served through
nginx, which would make `no_serve_uploads` entirely nonsensical of a
name.
Serve the files through Django, with an offload for the actual image
response to an internal nginx route. In development, serve the files
directly in Django.
We do _not_ mark the contents as immutable for caching purposes, since
the path for avatar images is hashed only by their user-id and a salt,
and as such are reused when a user's avatar is updated.
The `django-sendfile2` module unfortunately only supports a single
`SENDFILE` root path -- an invariant which subsequent commits need to
break. Especially as Zulip only runs with a single webserver, and
thus sendfile backend, the functionality is simple to inline.
It is worth noting that the following headers from the initial Django
response are _preserved_, if present, and sent unmodified to the
client; all other headers are overridden by those supplied by the
internal redirect[^1]:
- Content-Type
- Content-Disposition
- Accept-Ranges
- Set-Cookie
- Cache-Control
- Expires
As such, we explicitly unset the Content-type header to allow nginx to
set it from the static file, but set Content-Disposition and
Cache-Control as we want them to be.
[^1]: https://www.nginx.com/resources/wiki/start/topics/examples/xsendfile/
sendfile already applied a Content-Disposition header, but the
algorithm may provide both `filename=` and `filename*=` values (which
is potentially confusing to clients) and incorrectly slash-escapes
quotes in Unicode strings.
Django provides a correct implementation, but it is only accessible to
FileResponse objects. Since the entire point is to offload the
filehandle handling, we cannot use a FileResponse.
Django 4.2 will make the function available outside of FileResponse.
Until then, extract our own Content-Disposition handling, based on
Django's.
We remove the very verbose comment added in d4360e2287, describing
Content-Disposition headers, as it does not add much.
Alex Vandiver
Mateusz Mandera
Anders Kaseorg
Sahil Batra
arghyadeep10
Ujjawal Modi
Lauryn Menard
AcKindle3