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.
This `mimetype` parameter was introduced in c4fa29a and its last
usage removed in 5bab2a3. This parameter was undocumented in the
OpenAPI endpoint documentation for `/user_uploads`, therefore
there shouldn't be client implementations that rely on it's
presence.
Removes the `request.GET` call for the `mimetype` parameter and
replaces it by getting the `content_type` value from the file,
which is an instance of Django's `UploadedFile` class and stores
that file metadata as a property.
If that returns `None` or an empty string, then we try to guess
the `content_type` from the filename, which is the same as the
previous behaviour when `mimetype` was `None` (which we assume
has been true since it's usage was removed; see above).
If unable to guess the `content_type` from the filename, we now
fallback to "application/octet-stream", instead of an empty string
or `None` value.
Also, removes the specific test written for having `mimetype` as
a url parameter in the request, and replaces it with a test that
covers when we try to guess `content_type` from the filename.
We have already checked the size of the file in `upload_file_backend`.
This is the only caller of `upload_message_image_from_request`, and
indirectly the only caller of `get_file_info`. There is no need to
retrieve this information again.
Signed-off-by: Zixuan James Li <p359101898@gmail.com>
Adds request as a parameter to json_success as a refactor towards
making `ignored_parameters_unsupported` functionality available
for all API endpoints.
Also, removes any data parameters that are an empty dict or
a dict with the generic success response values.
JsonableError has two major benefits over json_error:
* It can be raised from anywhere in the codebase, rather than
being a return value, which is much more convenient for refactoring,
as one doesn't potentially need to change error handling style when
extracting a bit of view code to a function.
* It is guaranteed to contain the `code` property, which is helpful
for API consistency.
Various stragglers are not updated because JsonableError requires
subclassing in order to specify custom data or HTTP status codes.
django.utils.translation.ugettext is a deprecated alias of
django.utils.translation.gettext as of Django 3.0, and will be removed
in Django 4.0.
Signed-off-by: Anders Kaseorg <anders@zulip.com>
Fixes#2665.
Regenerated by tabbott with `lint --fix` after a rebase and change in
parameters.
Note from tabbott: In a few cases, this converts technical debt in the
form of unsorted imports into different technical debt in the form of
our largest files having very long, ugly import sequences at the
start. I expect this change will increase pressure for us to split
those files, which isn't a bad thing.
Signed-off-by: Anders Kaseorg <anders@zulip.com>
This is be useful for the mobile and desktop apps to hand an uploaded
file off to the system browser so that it can render PDFs (Etc.).
The S3 backend implementation is simple; for the local upload backend,
we use Django's signing feature to simulate the same sort of 60-second
lifetime token.
Co-Author-By: Mateusz Mandera <mateusz.mandera@protonmail.com>
Django 2.2.x is the next LTS release after Django 1.11.x; I expect
we'll be on it for a while, as Django 3.x won't have an LTS release
series out for a while.
Because of upstream API changes in Django, this commit includes
several changes beyond requirements and:
* urls: django.urls.resolvers.RegexURLPattern has been replaced by
django.urls.resolvers.URLPattern; affects OpenAPI code and related
features which re-parse Django's internals.
https://code.djangoproject.com/ticket/28593
* test_runner: Change number to suffix. Django changed the name in this
ticket: https://code.djangoproject.com/ticket/28578
* Delete now-unnecessary SameSite cookie code (it's now the default).
* forms: urlsafe_base64_encode returns string in Django 2.2.
https://docs.djangoproject.com/en/2.2/ref/utils/#django.utils.http.urlsafe_base64_encode
* upload: Django's File.size property replaces _get_size().
https://docs.djangoproject.com/en/2.2/_modules/django/core/files/base/
* process_queue: Migrate to new autoreload API.
* test_messages: Add an extra query caused by .refresh_from_db() losing
the .select_related() on the Realm object.
* session: Sync SessionHostDomainMiddleware with Django 2.2.
There's a lot more we can do to take advantage of the new release;
this is tracked in #11341.
Many changes by Tim Abbott, Umair Waheed, and Mateusz Mandera squashed
are squashed into this commit.
Fixes#10835.
Apparently, our change in b8a1050fc4 to
stop caching responses on API endpoints accidentally ended up
affecting uploaded files as well.
Fix this by explicitly setting a Cache-Control header in our Sendfile
responses, as well as changing our outer API caching code to only set
the never cache headers if the view function didn't explicitly specify
them itself.
This is not directly related to #13088, as that is a similar issue
with the S3 backend.
Thanks to Gert Burger for the report.
* Whitelist a small number of image/ types to be served as
non-attachments.
* Serve the file using the type that we validated rather than relying
on an independent guess to match.
This issue can lead to a stored XSS security vulnerability for older
browsers that don't support Content-Security-Policy.
It primarily affects servers using Zulip's local file uploads backend
for servers running Ubuntu 16.04 Xenial or newer; the legacy local
file upload backend for (now EOL) Ubuntu 14.04 Trusty was not affected
and it has limited impact for the S3 upload backend (which uses an
unprivileged S3 bucket domain to serve files).
Signed-off-by: Anders Kaseorg <anders@zulipchat.com>
The original seems to be unmaintained
(johnsensible/django-sendfile#65). Notably, this fixes a bug in the
filename parameter, which perviously showed the Python 3 repr of a
byte string (johnsensible/django-sendfile#49).
Signed-off-by: Anders Kaseorg <anders@zulipchat.com>
This reverts commit fd9dd51d16 (#1815).
The issue described does not exist in Python 3, where urllib.parse now
_only_ accepts (Unicode) str and does the right thing with it. The
workaround was not being triggered and would have failed if it were.
Signed-off-by: Anders Kaseorg <anders@zulipchat.com>
We start to force downloads for the attachment files. We do this
for all files except images or pdf's. We would like images or pdf's
to open up in browser itself.
Tweaked by tabbott for comment clarity and correctness.
From here on we start to authenticate uploaded file request before
serving this files in production. This involves allowing NGINX to
pass on these file requests to Django for authentication and then
serve these files by making use on internal redirect requests having
x-accel-redirect field. The redirection on requests and loading
of x-accel-redirect param is handled by django-sendfile.
NOTE: This commit starts to authenticate these requests for Zulip
servers running platforms either Ubuntu Xenial (16.04) or above.
Fixes: #320 and #291 partially.
We'll replace this primarily with per-realm quotas (plus the simple
per-file limit of settings.MAX_FILE_UPLOAD_SIZE, 25 MiB by default).
We do want per-user quotas too, but they'll need some more management
apparatus around them so an admin has a practical way to set them
differently for different users. And the error handling in this
existing code is rather confused. Just clear this feature out
entirely for now; then we'll build the per-realm version more cleanly,
and then we can later add back per-realm quotas modelled after that.
The migration to actually remove the field is in a subsequent commit.
Based in part on work by Vishnu Ks (hackerkid).
This is a remerge of e985b57259 (after
resolving merge conflicts, updating the tests, adding mypy annotations
etc.), which should now be correct, because we've done the necessary
database migration.
The rebase/remerge work was done by Tim Abbott and Aditya Bansal.
This is an important part of #320.
Anders Kaseorg
Alex Vandiver
Lauryn Menard
Zixuan James Li
Aman Agrawal
Lauryn Menard
PIG208
Mateusz Mandera
Tim Abbott
rht
Anders Kaseorg
Anders Kaseorg
Aditya Bansal
Vishnu Ks
Greg Price
Ethan
rahuldeve
Rishi Gupta
Philip Skomorokhov
Yago González