discourse/lib/s3_helper.rb

Ignoring revisions in .git-blame-ignore-revs. Click here to bypass and see the normal blame view.

447 lines
12 KiB
Ruby
Raw Normal View History

# frozen_string_literal: true
require "aws-sdk-s3"
class S3Helper
FIFTEEN_MEGABYTES = 15 * 1024 * 1024
class SettingMissing < StandardError; end
attr_reader :s3_bucket_name, :s3_bucket_folder_path
##
# Controls the following:
#
# * cache time for secure-media URLs
# * expiry time for S3 presigned URLs, which include backup downloads and
# any upload that has a private ACL (e.g. secure uploads)
#
# SiteSetting.s3_presigned_get_url_expires_after_seconds
FEATURE: Initial implementation of direct S3 uploads with uppy and stubs (#13787) This adds a few different things to allow for direct S3 uploads using uppy. **These changes are still not the default.** There are hidden `enable_experimental_image_uploader` and `enable_direct_s3_uploads` settings that must be turned on for any of this code to be used, and even if they are turned on only the User Card Background for the user profile actually uses uppy-image-uploader. A new `ExternalUploadStub` model and database table is introduced in this pull request. This is used to keep track of uploads that are uploaded to a temporary location in S3 with the direct to S3 code, and they are eventually deleted a) when the direct upload is completed and b) after a certain time period of not being used. ### Starting a direct S3 upload When an S3 direct upload is initiated with uppy, we first request a presigned PUT URL from the new `generate-presigned-put` endpoint in `UploadsController`. This generates an S3 key in the `temp` folder inside the correct bucket path, along with any metadata from the clientside (e.g. the SHA1 checksum described below). This will also create an `ExternalUploadStub` and store the details of the temp object key and the file being uploaded. Once the clientside has this URL, uppy will upload the file direct to S3 using the presigned URL. Once the upload is complete we go to the next stage. ### Completing a direct S3 upload Once the upload to S3 is done we call the new `complete-external-upload` route with the unique identifier of the `ExternalUploadStub` created earlier. Only the user who made the stub can complete the external upload. One of two paths is followed via the `ExternalUploadManager`. 1. If the object in S3 is too large (currently 100mb defined by `ExternalUploadManager::DOWNLOAD_LIMIT`) we do not download and generate the SHA1 for that file. Instead we create the `Upload` record via `UploadCreator` and simply copy it to its final destination on S3 then delete the initial temp file. Several modifications to `UploadCreator` have been made to accommodate this. 2. If the object in S3 is small enough, we download it. When the temporary S3 file is downloaded, we compare the SHA1 checksum generated by the browser with the actual SHA1 checksum of the file generated by ruby. The browser SHA1 checksum is stored on the object in S3 with metadata, and is generated via the `UppyChecksum` plugin. Keep in mind that some browsers will not generate this due to compatibility or other issues. We then follow the normal `UploadCreator` path with one exception. To cut down on having to re-upload the file again, if there are no changes (such as resizing etc) to the file in `UploadCreator` we follow the same copy + delete temp path that we do for files that are too large. 3. Finally we return the serialized upload record back to the client There are several errors that could happen that are handled by `UploadsController` as well. Also in this PR is some refactoring of `displayErrorForUpload` to handle both uppy and jquery file uploader errors.
2021-07-27 18:42:25 -04:00
##
# Controls the following:
#
# * presigned put_object URLs for direct S3 uploads
UPLOAD_URL_EXPIRES_AFTER_SECONDS ||= 10.minutes.to_i
def initialize(s3_bucket_name, tombstone_prefix = '', options = {})
@s3_client = options.delete(:client)
@s3_options = default_s3_options.merge(options)
@s3_bucket_name, @s3_bucket_folder_path = begin
raise Discourse::InvalidParameters.new("s3_bucket_name") if s3_bucket_name.blank?
self.class.get_bucket_and_folder_path(s3_bucket_name)
end
@tombstone_prefix =
if @s3_bucket_folder_path
File.join(@s3_bucket_folder_path, tombstone_prefix)
else
tombstone_prefix
end
end
def self.build_from_config(use_db_s3_config: false, for_backup: false, s3_client: nil)
setting_klass = use_db_s3_config ? SiteSetting : GlobalSetting
options = S3Helper.s3_options(setting_klass)
options[:client] = s3_client if s3_client.present?
bucket =
if for_backup
setting_klass.s3_backup_bucket
else
use_db_s3_config ? SiteSetting.s3_upload_bucket : GlobalSetting.s3_bucket
end
S3Helper.new(bucket.downcase, '', options)
end
def self.get_bucket_and_folder_path(s3_bucket_name)
s3_bucket_name.downcase.split("/", 2)
end
def upload(file, path, options = {})
path = get_path_for_s3_upload(path)
obj = s3_bucket.object(path)
etag = begin
if File.size(file.path) >= FIFTEEN_MEGABYTES
options[:multipart_threshold] = FIFTEEN_MEGABYTES
obj.upload_file(file, options)
obj.load
obj.etag
else
options[:body] = file
obj.put(options).etag
end
end
[path, etag.gsub('"', '')]
end
FEATURE: Direct S3 multipart uploads for backups (#14736) This PR introduces a new `enable_experimental_backup_uploads` site setting (default false and hidden), which when enabled alongside `enable_direct_s3_uploads` will allow for direct S3 multipart uploads of backup .tar.gz files. To make multipart external uploads work with both the S3BackupStore and the S3Store, I've had to move several methods out of S3Store and into S3Helper, including: * presigned_url * create_multipart * abort_multipart * complete_multipart * presign_multipart_part * list_multipart_parts Then, S3Store and S3BackupStore either delegate directly to S3Helper or have their own special methods to call S3Helper for these methods. FileStore.temporary_upload_path has also removed its dependence on upload_path, and can now be used interchangeably between the stores. A similar change was made in the frontend as well, moving the multipart related JS code out of ComposerUppyUpload and into a mixin of its own, so it can also be used by UppyUploadMixin. Some changes to ExternalUploadManager had to be made here as well. The backup direct uploads do not need an Upload record made for them in the database, so they can be moved to their final S3 resting place when completing the multipart upload. This changeset is not perfect; it introduces some special cases in UploadController to handle backups that was previously in BackupController, because UploadController is where the multipart routes are located. A subsequent pull request will pull these routes into a module or some other sharing pattern, along with hooks, so the backup controller and the upload controller (and any future controllers that may need them) can include these routes in a nicer way.
2021-11-10 17:25:31 -05:00
def path_from_url(url)
URI.parse(url).path.delete_prefix("/")
end
def remove(s3_filename, copy_to_tombstone = false)
s3_filename = s3_filename.dup
# copy the file in tombstone
if copy_to_tombstone && @tombstone_prefix.present?
2018-08-07 23:26:05 -04:00
self.copy(
get_path_for_s3_upload(s3_filename),
File.join(@tombstone_prefix, s3_filename)
2018-08-07 23:26:05 -04:00
)
end
# delete the file
s3_filename.prepend(multisite_upload_path) if Rails.configuration.multisite
delete_object(get_path_for_s3_upload(s3_filename))
rescue Aws::S3::Errors::NoSuchKey, Aws::S3::Errors::NotFound
end
def delete_object(key)
s3_bucket.object(key).delete
rescue Aws::S3::Errors::NoSuchKey, Aws::S3::Errors::NotFound
end
def delete_objects(keys)
s3_bucket.delete_objects({
delete: {
objects: keys.map { |k| { key: k } },
quiet: true,
},
})
end
def copy(source, destination, options: {})
if options[:apply_metadata_to_destination]
options = options.except(:apply_metadata_to_destination).merge(metadata_directive: "REPLACE")
end
FEATURE: Initial implementation of direct S3 uploads with uppy and stubs (#13787) This adds a few different things to allow for direct S3 uploads using uppy. **These changes are still not the default.** There are hidden `enable_experimental_image_uploader` and `enable_direct_s3_uploads` settings that must be turned on for any of this code to be used, and even if they are turned on only the User Card Background for the user profile actually uses uppy-image-uploader. A new `ExternalUploadStub` model and database table is introduced in this pull request. This is used to keep track of uploads that are uploaded to a temporary location in S3 with the direct to S3 code, and they are eventually deleted a) when the direct upload is completed and b) after a certain time period of not being used. ### Starting a direct S3 upload When an S3 direct upload is initiated with uppy, we first request a presigned PUT URL from the new `generate-presigned-put` endpoint in `UploadsController`. This generates an S3 key in the `temp` folder inside the correct bucket path, along with any metadata from the clientside (e.g. the SHA1 checksum described below). This will also create an `ExternalUploadStub` and store the details of the temp object key and the file being uploaded. Once the clientside has this URL, uppy will upload the file direct to S3 using the presigned URL. Once the upload is complete we go to the next stage. ### Completing a direct S3 upload Once the upload to S3 is done we call the new `complete-external-upload` route with the unique identifier of the `ExternalUploadStub` created earlier. Only the user who made the stub can complete the external upload. One of two paths is followed via the `ExternalUploadManager`. 1. If the object in S3 is too large (currently 100mb defined by `ExternalUploadManager::DOWNLOAD_LIMIT`) we do not download and generate the SHA1 for that file. Instead we create the `Upload` record via `UploadCreator` and simply copy it to its final destination on S3 then delete the initial temp file. Several modifications to `UploadCreator` have been made to accommodate this. 2. If the object in S3 is small enough, we download it. When the temporary S3 file is downloaded, we compare the SHA1 checksum generated by the browser with the actual SHA1 checksum of the file generated by ruby. The browser SHA1 checksum is stored on the object in S3 with metadata, and is generated via the `UppyChecksum` plugin. Keep in mind that some browsers will not generate this due to compatibility or other issues. We then follow the normal `UploadCreator` path with one exception. To cut down on having to re-upload the file again, if there are no changes (such as resizing etc) to the file in `UploadCreator` we follow the same copy + delete temp path that we do for files that are too large. 3. Finally we return the serialized upload record back to the client There are several errors that could happen that are handled by `UploadsController` as well. Also in this PR is some refactoring of `displayErrorForUpload` to handle both uppy and jquery file uploader errors.
2021-07-27 18:42:25 -04:00
destination = get_path_for_s3_upload(destination)
source_object = if !Rails.configuration.multisite || source.include?(multisite_upload_path) || source.include?(@tombstone_prefix)
s3_bucket.object(source)
elsif @s3_bucket_folder_path
folder, filename = source.split("/", 2)
s3_bucket.object(File.join(folder, multisite_upload_path, filename))
else
s3_bucket.object(File.join(multisite_upload_path, source))
end
if source_object.size > FIFTEEN_MEGABYTES
options[:multipart_copy] = true
options[:content_length] = source_object.size
end
FEATURE: Initial implementation of direct S3 uploads with uppy and stubs (#13787) This adds a few different things to allow for direct S3 uploads using uppy. **These changes are still not the default.** There are hidden `enable_experimental_image_uploader` and `enable_direct_s3_uploads` settings that must be turned on for any of this code to be used, and even if they are turned on only the User Card Background for the user profile actually uses uppy-image-uploader. A new `ExternalUploadStub` model and database table is introduced in this pull request. This is used to keep track of uploads that are uploaded to a temporary location in S3 with the direct to S3 code, and they are eventually deleted a) when the direct upload is completed and b) after a certain time period of not being used. ### Starting a direct S3 upload When an S3 direct upload is initiated with uppy, we first request a presigned PUT URL from the new `generate-presigned-put` endpoint in `UploadsController`. This generates an S3 key in the `temp` folder inside the correct bucket path, along with any metadata from the clientside (e.g. the SHA1 checksum described below). This will also create an `ExternalUploadStub` and store the details of the temp object key and the file being uploaded. Once the clientside has this URL, uppy will upload the file direct to S3 using the presigned URL. Once the upload is complete we go to the next stage. ### Completing a direct S3 upload Once the upload to S3 is done we call the new `complete-external-upload` route with the unique identifier of the `ExternalUploadStub` created earlier. Only the user who made the stub can complete the external upload. One of two paths is followed via the `ExternalUploadManager`. 1. If the object in S3 is too large (currently 100mb defined by `ExternalUploadManager::DOWNLOAD_LIMIT`) we do not download and generate the SHA1 for that file. Instead we create the `Upload` record via `UploadCreator` and simply copy it to its final destination on S3 then delete the initial temp file. Several modifications to `UploadCreator` have been made to accommodate this. 2. If the object in S3 is small enough, we download it. When the temporary S3 file is downloaded, we compare the SHA1 checksum generated by the browser with the actual SHA1 checksum of the file generated by ruby. The browser SHA1 checksum is stored on the object in S3 with metadata, and is generated via the `UppyChecksum` plugin. Keep in mind that some browsers will not generate this due to compatibility or other issues. We then follow the normal `UploadCreator` path with one exception. To cut down on having to re-upload the file again, if there are no changes (such as resizing etc) to the file in `UploadCreator` we follow the same copy + delete temp path that we do for files that are too large. 3. Finally we return the serialized upload record back to the client There are several errors that could happen that are handled by `UploadsController` as well. Also in this PR is some refactoring of `displayErrorForUpload` to handle both uppy and jquery file uploader errors.
2021-07-27 18:42:25 -04:00
destination_object = s3_bucket.object(destination)
# Note for small files that do not use multipart copy: Any options for metadata
# (e.g. content_disposition, content_type) will not be applied unless the
# metadata_directive = "REPLACE" option is passed in. If this is not passed in,
# the source object's metadata will be used.
# For larger files it copies the metadata from the source file and merges it
# with values from the copy call.
response = destination_object.copy_from(source_object, options)
etag = if response.respond_to?(:copy_object_result)
# small files, regular copy
response.copy_object_result.etag
else
# larger files, multipart copy
response.data.etag
end
[destination, etag.gsub('"', '')]
2018-08-07 23:26:05 -04:00
end
# Several places in the application need certain CORS rules to exist
# inside an S3 bucket so requests to the bucket can be made
# directly from the browser. The s3:ensure_cors_rules rake task
# is used to ensure these rules exist for assets, S3 backups, and
# direct S3 uploads, depending on configuration.
def ensure_cors!(rules = nil)
return unless SiteSetting.s3_install_cors_rule
rules = [rules] if !rules.is_a?(Array)
existing_rules = fetch_bucket_cors_rules
new_rules = rules - existing_rules
return false if new_rules.empty?
2017-10-08 19:26:58 -04:00
final_rules = existing_rules + new_rules
2017-10-08 19:26:58 -04:00
begin
2017-10-08 19:26:58 -04:00
s3_resource.client.put_bucket_cors(
bucket: @s3_bucket_name,
cors_configuration: {
cors_rules: final_rules
2017-10-08 19:26:58 -04:00
}
)
rescue Aws::S3::Errors::AccessDenied => err
# TODO (martin) Remove this warning log level once we are sure this new
# ensure_cors! rule is functioning correctly.
Discourse.warn_exception(err, message: "Could not PutBucketCors rules for #{@s3_bucket_name}, rules: #{final_rules}")
return false
2017-10-08 19:26:58 -04:00
end
true
2017-10-08 19:26:58 -04:00
end
def update_lifecycle(id, days, prefix: nil, tag: nil)
filter = {}
if prefix
filter[:prefix] = prefix
elsif tag
filter[:tag] = tag
end
# cf. http://docs.aws.amazon.com/AmazonS3/latest/dev/object-lifecycle-mgmt.html
rule = {
id: id,
status: "Enabled",
2017-10-08 19:26:58 -04:00
expiration: { days: days },
filter: filter
}
2017-10-08 19:26:58 -04:00
rules = []
2017-10-08 19:26:58 -04:00
begin
rules = s3_resource.client.get_bucket_lifecycle_configuration(bucket: @s3_bucket_name).rules
rescue Aws::S3::Errors::NoSuchLifecycleConfiguration
# skip trying to merge
end
# in the past we has a rule that was called purge-tombstone vs purge_tombstone
# just go ahead and normalize for our bucket
rules.delete_if do |r|
r.id.gsub('_', '-') == id.gsub('_', '-')
end
rules << rule
# normalize filter in rules, due to AWS library bug
rules = rules.map do |r|
r = r.to_h
prefix = r.delete(:prefix)
if prefix
r[:filter] = { prefix: prefix }
end
r
end
2017-10-08 19:26:58 -04:00
s3_resource.client.put_bucket_lifecycle_configuration(
bucket: @s3_bucket_name,
lifecycle_configuration: {
rules: rules
})
end
def update_tombstone_lifecycle(grace_period)
return if !SiteSetting.s3_configure_tombstone_policy
return if @tombstone_prefix.blank?
update_lifecycle("purge_tombstone", grace_period, prefix: @tombstone_prefix)
end
def list(prefix = "", marker = nil)
options = { prefix: get_path_for_s3_upload(prefix) }
options[:marker] = marker if marker.present?
s3_bucket.objects(options)
end
def tag_file(key, tags)
tag_array = []
tags.each do |k, v|
tag_array << { key: k.to_s, value: v.to_s }
end
s3_resource.client.put_object_tagging(
bucket: @s3_bucket_name,
key: key,
tagging: {
tag_set: tag_array
}
)
end
def object(path)
2018-12-26 11:34:49 -05:00
s3_bucket.object(get_path_for_s3_upload(path))
end
def self.s3_options(obj)
2018-12-26 11:34:49 -05:00
opts = {
region: obj.s3_region
2018-12-26 11:34:49 -05:00
}
opts[:endpoint] = SiteSetting.s3_endpoint if SiteSetting.s3_endpoint.present?
opts[:http_continue_timeout] = SiteSetting.s3_http_continue_timeout
unless obj.s3_use_iam_profile
opts[:access_key_id] = obj.s3_access_key_id
opts[:secret_access_key] = obj.s3_secret_access_key
end
opts
end
def download_file(filename, destination_path, failure_message = nil)
object(filename).download_file(destination_path)
rescue => err
raise failure_message&.to_s || "Failed to download #{filename} because #{err.message.length > 0 ? err.message : err.class.to_s}"
end
def s3_client
@s3_client ||= Aws::S3::Client.new(@s3_options)
end
def s3_inventory_path(path = 'inventory')
get_path_for_s3_upload(path)
end
FEATURE: Direct S3 multipart uploads for backups (#14736) This PR introduces a new `enable_experimental_backup_uploads` site setting (default false and hidden), which when enabled alongside `enable_direct_s3_uploads` will allow for direct S3 multipart uploads of backup .tar.gz files. To make multipart external uploads work with both the S3BackupStore and the S3Store, I've had to move several methods out of S3Store and into S3Helper, including: * presigned_url * create_multipart * abort_multipart * complete_multipart * presign_multipart_part * list_multipart_parts Then, S3Store and S3BackupStore either delegate directly to S3Helper or have their own special methods to call S3Helper for these methods. FileStore.temporary_upload_path has also removed its dependence on upload_path, and can now be used interchangeably between the stores. A similar change was made in the frontend as well, moving the multipart related JS code out of ComposerUppyUpload and into a mixin of its own, so it can also be used by UppyUploadMixin. Some changes to ExternalUploadManager had to be made here as well. The backup direct uploads do not need an Upload record made for them in the database, so they can be moved to their final S3 resting place when completing the multipart upload. This changeset is not perfect; it introduces some special cases in UploadController to handle backups that was previously in BackupController, because UploadController is where the multipart routes are located. A subsequent pull request will pull these routes into a module or some other sharing pattern, along with hooks, so the backup controller and the upload controller (and any future controllers that may need them) can include these routes in a nicer way.
2021-11-10 17:25:31 -05:00
def abort_multipart(key:, upload_id:)
s3_client.abort_multipart_upload(
bucket: s3_bucket_name,
key: key,
upload_id: upload_id
)
end
def create_multipart(key, content_type, metadata: {})
response = s3_client.create_multipart_upload(
acl: "private",
bucket: s3_bucket_name,
key: key,
content_type: content_type,
metadata: metadata
)
{ upload_id: response.upload_id, key: key }
end
def presign_multipart_part(upload_id:, key:, part_number:)
presigned_url(
key,
method: :upload_part,
expires_in: S3Helper::UPLOAD_URL_EXPIRES_AFTER_SECONDS,
opts: {
part_number: part_number,
upload_id: upload_id
}
)
end
# Important note from the S3 documentation:
#
# This request returns a default and maximum of 1000 parts.
# You can restrict the number of parts returned by specifying the
# max_parts argument. If your multipart upload consists of more than 1,000
# parts, the response returns an IsTruncated field with the value of true,
# and a NextPartNumberMarker element.
#
# In subsequent ListParts requests you can include the part_number_marker arg
# using the NextPartNumberMarker the field value from the previous response to
# get more parts.
#
# See https://docs.aws.amazon.com/sdk-for-ruby/v3/api/Aws/S3/Client.html#list_parts-instance_method
def list_multipart_parts(upload_id:, key:, max_parts: 1000, start_from_part_number: nil)
options = {
bucket: s3_bucket_name,
key: key,
upload_id: upload_id,
max_parts: max_parts
}
if start_from_part_number.present?
options[:part_number_marker] = start_from_part_number
end
s3_client.list_parts(options)
end
def complete_multipart(upload_id:, key:, parts:)
s3_client.complete_multipart_upload(
bucket: s3_bucket_name,
key: key,
upload_id: upload_id,
multipart_upload: {
parts: parts
}
)
end
def presigned_url(
key,
method:,
expires_in: S3Helper::UPLOAD_URL_EXPIRES_AFTER_SECONDS,
opts: {}
)
Aws::S3::Presigner.new(client: s3_client).presigned_url(
method,
{
bucket: s3_bucket_name,
key: key,
expires_in: expires_in,
}.merge(opts)
)
end
private
def fetch_bucket_cors_rules
begin
s3_resource.client.get_bucket_cors(
bucket: @s3_bucket_name
).cors_rules&.map(&:to_h) || []
rescue Aws::S3::Errors::NoSuchCORSConfiguration
# no rule
[]
end
end
def default_s3_options
if SiteSetting.enable_s3_uploads?
options = self.class.s3_options(SiteSetting)
check_missing_site_options
options
elsif GlobalSetting.use_s3?
self.class.s3_options(GlobalSetting)
else
{}
end
end
def get_path_for_s3_upload(path)
if @s3_bucket_folder_path &&
!path.starts_with?(@s3_bucket_folder_path) &&
!path.starts_with?(File.join(FileStore::BaseStore::TEMPORARY_UPLOAD_PREFIX, @s3_bucket_folder_path))
return File.join(@s3_bucket_folder_path, path)
end
path
end
def multisite_upload_path
path = File.join("uploads", RailsMultisite::ConnectionManagement.current_db, "/")
return path if !Rails.env.test?
File.join(path, "test_#{ENV['TEST_ENV_NUMBER'].presence || '0'}", "/")
end
def s3_resource
Aws::S3::Resource.new(client: s3_client)
end
def s3_bucket
2018-08-07 23:26:05 -04:00
@s3_bucket ||= begin
bucket = s3_resource.bucket(@s3_bucket_name)
bucket.create unless bucket.exists?
bucket
end
end
def check_missing_site_options
unless SiteSetting.s3_use_iam_profile
raise SettingMissing.new("access_key_id") if SiteSetting.s3_access_key_id.blank?
raise SettingMissing.new("secret_access_key") if SiteSetting.s3_secret_access_key.blank?
end
end
end