PeerTube/shared/server-commands/server/server.ts

438 lines
13 KiB
TypeScript
Raw Normal View History

2021-07-16 02:47:51 -05:00
import { ChildProcess, fork } from 'child_process'
import { copy } from 'fs-extra'
import { join } from 'path'
import { parallelTests, randomInt, root } from '@shared/core-utils'
import { Video, VideoChannel, VideoChannelSync, VideoCreateResult, VideoDetails } from '@shared/models'
2021-07-16 02:47:51 -05:00
import { BulkCommand } from '../bulk'
import { CLICommand } from '../cli'
import { CustomPagesCommand } from '../custom-pages'
import { FeedCommand } from '../feeds'
import { LogsCommand } from '../logs'
import { SQLCommand } from '../miscs'
2021-07-16 02:47:51 -05:00
import { AbusesCommand } from '../moderation'
import { OverviewsCommand } from '../overviews'
import { SearchCommand } from '../search'
import { SocketIOCommand } from '../socket'
import {
AccountsCommand,
BlocklistCommand,
LoginCommand,
NotificationsCommand,
2023-01-19 02:28:29 -06:00
RegistrationsCommand,
SubscriptionsCommand,
TwoFactorCommand,
UsersCommand
} from '../users'
2021-07-16 02:47:51 -05:00
import {
BlacklistCommand,
CaptionsCommand,
ChangeOwnershipCommand,
ChannelsCommand,
Channel sync (#5135) * Add external channel URL for channel update / creation (#754) * Disallow synchronisation if user has no video quota (#754) * More constraints serverside (#754) * Disable sync if server configuration does not allow HTTP import (#754) * Working version synchronizing videos with a job (#754) TODO: refactoring, too much code duplication * More logs and try/catch (#754) * Fix eslint error (#754) * WIP: support synchronization time change (#754) * New frontend #754 * WIP: Create sync front (#754) * Enhance UI, sync creation form (#754) * Warning message when HTTP upload is disallowed * More consistent names (#754) * Binding Front with API (#754) * Add a /me API (#754) * Improve list UI (#754) * Implement creation and deletion routes (#754) * Lint (#754) * Lint again (#754) * WIP: UI for triggering import existing videos (#754) * Implement jobs for syncing and importing channels * Don't sync videos before sync creation + avoid concurrency issue (#754) * Cleanup (#754) * Cleanup: OpenAPI + API rework (#754) * Remove dead code (#754) * Eslint (#754) * Revert the mess with whitespaces in constants.ts (#754) * Some fixes after rebase (#754) * Several fixes after PR remarks (#754) * Front + API: Rename video-channels-sync to video-channel-syncs (#754) * Allow enabling channel sync through UI (#754) * getChannelInfo (#754) * Minor fixes: openapi + model + sql (#754) * Simplified API validators (#754) * Rename MChannelSync to MChannelSyncChannel (#754) * Add command for VideoChannelSync (#754) * Use synchronization.enabled config (#754) * Check parameters test + some fixes (#754) * Fix conflict mistake (#754) * Restrict access to video channel sync list API (#754) * Start adding unit test for synchronization (#754) * Continue testing (#754) * Tests finished + convertion of job to scheduler (#754) * Add lastSyncAt field (#754) * Fix externalRemoteUrl sort + creation date not well formatted (#754) * Small fix (#754) * Factorize addYoutubeDLImport and buildVideo (#754) * Check duplicates on channel not on users (#754) * factorize thumbnail generation (#754) * Fetch error should return status 400 (#754) * Separate video-channel-import and video-channel-sync-latest (#754) * Bump DB migration version after rebase (#754) * Prettier states in UI table (#754) * Add DefaultScope in VideoChannelSyncModel (#754) * Fix audit logs (#754) * Ensure user can upload when importing channel + minor fixes (#754) * Mark synchronization as failed on exception + typos (#754) * Change REST API for importing videos into channel (#754) * Add option for fully synchronize a chnanel (#754) * Return a whole sync object on creation to avoid tricks in Front (#754) * Various remarks (#754) * Single quotes by default (#754) * Rename synchronization to video_channel_synchronization * Add check.latest_videos_count and max_per_user options (#754) * Better channel rendering in list #754 * Allow sorting with channel name and state (#754) * Add missing tests for channel imports (#754) * Prefer using a parent job for channel sync * Styling * Client styling Co-authored-by: Chocobozzz <me@florianbigard.com>
2022-08-10 02:53:39 -05:00
ChannelSyncsCommand,
2021-07-16 02:47:51 -05:00
HistoryCommand,
ImportsCommand,
LiveCommand,
PlaylistsCommand,
ServicesCommand,
StreamingPlaylistsCommand,
VideosCommand,
2022-03-22 10:58:49 -05:00
VideoStudioCommand,
VideoTokenCommand,
ViewsCommand
2021-07-16 02:47:51 -05:00
} from '../videos'
import { CommentsCommand } from '../videos/comments-command'
import { VideoStatsCommand } from '../videos/video-stats-command'
2021-07-16 02:47:51 -05:00
import { ConfigCommand } from './config-command'
import { ContactFormCommand } from './contact-form-command'
import { DebugCommand } from './debug-command'
import { FollowsCommand } from './follows-command'
import { JobsCommand } from './jobs-command'
import { MetricsCommand } from './metrics-command'
import { ObjectStorageCommand } from './object-storage-command'
2021-07-16 02:47:51 -05:00
import { PluginsCommand } from './plugins-command'
import { RedundancyCommand } from './redundancy-command'
import { ServersCommand } from './servers-command'
import { StatsCommand } from './stats-command'
export type RunServerOptions = {
hideLogs?: boolean
2021-07-22 05:07:26 -05:00
nodeArgs?: string[]
peertubeArgs?: string[]
Add support for saving video files to object storage (#4290) * Add support for saving video files to object storage * Add support for custom url generation on s3 stored files Uses two config keys to support url generation that doesn't directly go to (compatible s3). Can be used to generate urls to any cache server or CDN. * Upload files to s3 concurrently and delete originals afterwards * Only publish after move to object storage is complete * Use base url instead of url template * Fix mistyped config field * Add rudenmentary way to download before transcode * Implement Chocobozzz suggestions https://github.com/Chocobozzz/PeerTube/pull/4290#issuecomment-891670478 The remarks in question: Try to use objectStorage prefix instead of s3 prefix for your function/variables/config names Prefer to use a tree for the config: s3.streaming_playlists_bucket -> object_storage.streaming_playlists.bucket Use uppercase for config: S3.STREAMING_PLAYLISTS_BUCKETINFO.bucket -> OBJECT_STORAGE.STREAMING_PLAYLISTS.BUCKET (maybe BUCKET_NAME instead of BUCKET) I suggest to rename moveJobsRunning to pendingMovingJobs (or better, create a dedicated videoJobInfo table with a pendingMove & videoId columns so we could also use this table to track pending transcoding jobs) https://github.com/Chocobozzz/PeerTube/pull/4290/files#diff-3e26d41ca4bda1de8e1747af70ca2af642abcc1e9e0bfb94239ff2165acfbde5R19 uses a string instead of an integer I think we should store the origin object storage URL in fileUrl, without base_url injection. Instead, inject the base_url at "runtime" so admins can easily change this configuration without running a script to update DB URLs * Import correct function * Support multipart upload * Remove import of node 15.0 module stream/promises * Extend maximum upload job length Using the same value as for redundancy downloading seems logical * Use dynamic part size for really large uploads Also adds very small part size for local testing * Fix decreasePendingMove query * Resolve various PR comments * Move to object storage after optimize * Make upload size configurable and increase default * Prune webtorrent files that are stored in object storage * Move files after transcoding jobs * Fix federation * Add video path manager * Support move to external storage job in client * Fix live object storage tests Co-authored-by: Chocobozzz <me@florianbigard.com>
2021-08-17 01:26:20 -05:00
env?: { [ id: string ]: string }
2021-07-16 02:47:51 -05:00
}
export class PeerTubeServer {
app?: ChildProcess
url: string
host?: string
hostname?: string
port?: number
rtmpPort?: number
2021-11-05 05:36:03 -05:00
rtmpsPort?: number
2021-07-16 02:47:51 -05:00
parallel?: boolean
internalServerNumber: number
serverNumber?: number
customConfigFile?: string
store?: {
client?: {
id?: string
secret?: string
}
user?: {
username: string
password: string
email?: string
}
channel?: VideoChannel
videoChannelSync?: Partial<VideoChannelSync>
2021-07-16 02:47:51 -05:00
2021-07-22 07:28:03 -05:00
video?: Video
videoCreated?: VideoCreateResult
videoDetails?: VideoDetails
2021-07-16 02:47:51 -05:00
videos?: { id: number, uuid: string }[]
}
accessToken?: string
refreshToken?: string
bulk?: BulkCommand
cli?: CLICommand
customPage?: CustomPagesCommand
feed?: FeedCommand
logs?: LogsCommand
abuses?: AbusesCommand
overviews?: OverviewsCommand
search?: SearchCommand
contactForm?: ContactFormCommand
debug?: DebugCommand
follows?: FollowsCommand
jobs?: JobsCommand
metrics?: MetricsCommand
2021-07-16 02:47:51 -05:00
plugins?: PluginsCommand
redundancy?: RedundancyCommand
stats?: StatsCommand
config?: ConfigCommand
socketIO?: SocketIOCommand
accounts?: AccountsCommand
blocklist?: BlocklistCommand
subscriptions?: SubscriptionsCommand
live?: LiveCommand
services?: ServicesCommand
blacklist?: BlacklistCommand
captions?: CaptionsCommand
changeOwnership?: ChangeOwnershipCommand
playlists?: PlaylistsCommand
history?: HistoryCommand
imports?: ImportsCommand
Channel sync (#5135) * Add external channel URL for channel update / creation (#754) * Disallow synchronisation if user has no video quota (#754) * More constraints serverside (#754) * Disable sync if server configuration does not allow HTTP import (#754) * Working version synchronizing videos with a job (#754) TODO: refactoring, too much code duplication * More logs and try/catch (#754) * Fix eslint error (#754) * WIP: support synchronization time change (#754) * New frontend #754 * WIP: Create sync front (#754) * Enhance UI, sync creation form (#754) * Warning message when HTTP upload is disallowed * More consistent names (#754) * Binding Front with API (#754) * Add a /me API (#754) * Improve list UI (#754) * Implement creation and deletion routes (#754) * Lint (#754) * Lint again (#754) * WIP: UI for triggering import existing videos (#754) * Implement jobs for syncing and importing channels * Don't sync videos before sync creation + avoid concurrency issue (#754) * Cleanup (#754) * Cleanup: OpenAPI + API rework (#754) * Remove dead code (#754) * Eslint (#754) * Revert the mess with whitespaces in constants.ts (#754) * Some fixes after rebase (#754) * Several fixes after PR remarks (#754) * Front + API: Rename video-channels-sync to video-channel-syncs (#754) * Allow enabling channel sync through UI (#754) * getChannelInfo (#754) * Minor fixes: openapi + model + sql (#754) * Simplified API validators (#754) * Rename MChannelSync to MChannelSyncChannel (#754) * Add command for VideoChannelSync (#754) * Use synchronization.enabled config (#754) * Check parameters test + some fixes (#754) * Fix conflict mistake (#754) * Restrict access to video channel sync list API (#754) * Start adding unit test for synchronization (#754) * Continue testing (#754) * Tests finished + convertion of job to scheduler (#754) * Add lastSyncAt field (#754) * Fix externalRemoteUrl sort + creation date not well formatted (#754) * Small fix (#754) * Factorize addYoutubeDLImport and buildVideo (#754) * Check duplicates on channel not on users (#754) * factorize thumbnail generation (#754) * Fetch error should return status 400 (#754) * Separate video-channel-import and video-channel-sync-latest (#754) * Bump DB migration version after rebase (#754) * Prettier states in UI table (#754) * Add DefaultScope in VideoChannelSyncModel (#754) * Fix audit logs (#754) * Ensure user can upload when importing channel + minor fixes (#754) * Mark synchronization as failed on exception + typos (#754) * Change REST API for importing videos into channel (#754) * Add option for fully synchronize a chnanel (#754) * Return a whole sync object on creation to avoid tricks in Front (#754) * Various remarks (#754) * Single quotes by default (#754) * Rename synchronization to video_channel_synchronization * Add check.latest_videos_count and max_per_user options (#754) * Better channel rendering in list #754 * Allow sorting with channel name and state (#754) * Add missing tests for channel imports (#754) * Prefer using a parent job for channel sync * Styling * Client styling Co-authored-by: Chocobozzz <me@florianbigard.com>
2022-08-10 02:53:39 -05:00
channelSyncs?: ChannelSyncsCommand
2021-07-16 02:47:51 -05:00
streamingPlaylists?: StreamingPlaylistsCommand
channels?: ChannelsCommand
comments?: CommentsCommand
sql?: SQLCommand
notifications?: NotificationsCommand
servers?: ServersCommand
login?: LoginCommand
users?: UsersCommand
Add support for saving video files to object storage (#4290) * Add support for saving video files to object storage * Add support for custom url generation on s3 stored files Uses two config keys to support url generation that doesn't directly go to (compatible s3). Can be used to generate urls to any cache server or CDN. * Upload files to s3 concurrently and delete originals afterwards * Only publish after move to object storage is complete * Use base url instead of url template * Fix mistyped config field * Add rudenmentary way to download before transcode * Implement Chocobozzz suggestions https://github.com/Chocobozzz/PeerTube/pull/4290#issuecomment-891670478 The remarks in question: Try to use objectStorage prefix instead of s3 prefix for your function/variables/config names Prefer to use a tree for the config: s3.streaming_playlists_bucket -> object_storage.streaming_playlists.bucket Use uppercase for config: S3.STREAMING_PLAYLISTS_BUCKETINFO.bucket -> OBJECT_STORAGE.STREAMING_PLAYLISTS.BUCKET (maybe BUCKET_NAME instead of BUCKET) I suggest to rename moveJobsRunning to pendingMovingJobs (or better, create a dedicated videoJobInfo table with a pendingMove & videoId columns so we could also use this table to track pending transcoding jobs) https://github.com/Chocobozzz/PeerTube/pull/4290/files#diff-3e26d41ca4bda1de8e1747af70ca2af642abcc1e9e0bfb94239ff2165acfbde5R19 uses a string instead of an integer I think we should store the origin object storage URL in fileUrl, without base_url injection. Instead, inject the base_url at "runtime" so admins can easily change this configuration without running a script to update DB URLs * Import correct function * Support multipart upload * Remove import of node 15.0 module stream/promises * Extend maximum upload job length Using the same value as for redundancy downloading seems logical * Use dynamic part size for really large uploads Also adds very small part size for local testing * Fix decreasePendingMove query * Resolve various PR comments * Move to object storage after optimize * Make upload size configurable and increase default * Prune webtorrent files that are stored in object storage * Move files after transcoding jobs * Fix federation * Add video path manager * Support move to external storage job in client * Fix live object storage tests Co-authored-by: Chocobozzz <me@florianbigard.com>
2021-08-17 01:26:20 -05:00
objectStorage?: ObjectStorageCommand
2022-03-22 10:58:49 -05:00
videoStudio?: VideoStudioCommand
2021-07-16 02:47:51 -05:00
videos?: VideosCommand
videoStats?: VideoStatsCommand
views?: ViewsCommand
twoFactor?: TwoFactorCommand
videoToken?: VideoTokenCommand
2023-01-19 02:28:29 -06:00
registrations?: RegistrationsCommand
2021-07-16 02:47:51 -05:00
constructor (options: { serverNumber: number } | { url: string }) {
if ((options as any).url) {
this.setUrl((options as any).url)
} else {
this.setServerNumber((options as any).serverNumber)
}
this.store = {
client: {
id: null,
secret: null
},
user: {
username: null,
password: null
}
}
this.assignCommands()
}
setServerNumber (serverNumber: number) {
this.serverNumber = serverNumber
this.parallel = parallelTests()
this.internalServerNumber = this.parallel ? this.randomServer() : this.serverNumber
this.rtmpPort = this.parallel ? this.randomRTMP() : 1936
2021-11-05 05:36:03 -05:00
this.rtmpsPort = this.parallel ? this.randomRTMP() : 1937
2021-07-16 02:47:51 -05:00
this.port = 9000 + this.internalServerNumber
2022-12-09 04:14:47 -06:00
this.url = `http://127.0.0.1:${this.port}`
this.host = `127.0.0.1:${this.port}`
this.hostname = '127.0.0.1'
2021-07-16 02:47:51 -05:00
}
setUrl (url: string) {
const parsed = new URL(url)
this.url = url
this.host = parsed.host
this.hostname = parsed.hostname
this.port = parseInt(parsed.port)
}
getDirectoryPath (directoryName: string) {
const testDirectory = 'test' + this.internalServerNumber
return join(root(), testDirectory, directoryName)
}
2021-07-22 05:07:26 -05:00
async flushAndRun (configOverride?: Object, options: RunServerOptions = {}) {
2021-07-16 02:47:51 -05:00
await ServersCommand.flushTests(this.internalServerNumber)
2021-07-22 05:07:26 -05:00
return this.run(configOverride, options)
2021-07-16 02:47:51 -05:00
}
2021-07-22 05:07:26 -05:00
async run (configOverrideArg?: any, options: RunServerOptions = {}) {
2021-07-16 02:47:51 -05:00
// These actions are async so we need to be sure that they have both been done
const serverRunString = {
'HTTP server listening': false
}
const key = 'Database peertube_test' + this.internalServerNumber + ' is ready'
serverRunString[key] = false
const regexps = {
client_id: 'Client id: (.+)',
client_secret: 'Client secret: (.+)',
user_username: 'Username: (.+)',
user_password: 'User password: (.+)'
}
await this.assignCustomConfigFile()
const configOverride = this.buildConfigOverride()
if (configOverrideArg !== undefined) {
Object.assign(configOverride, configOverrideArg)
}
// Share the environment
const env = Object.create(process.env)
env['NODE_ENV'] = 'test'
env['NODE_APP_INSTANCE'] = this.internalServerNumber.toString()
env['NODE_CONFIG'] = JSON.stringify(configOverride)
Add support for saving video files to object storage (#4290) * Add support for saving video files to object storage * Add support for custom url generation on s3 stored files Uses two config keys to support url generation that doesn't directly go to (compatible s3). Can be used to generate urls to any cache server or CDN. * Upload files to s3 concurrently and delete originals afterwards * Only publish after move to object storage is complete * Use base url instead of url template * Fix mistyped config field * Add rudenmentary way to download before transcode * Implement Chocobozzz suggestions https://github.com/Chocobozzz/PeerTube/pull/4290#issuecomment-891670478 The remarks in question: Try to use objectStorage prefix instead of s3 prefix for your function/variables/config names Prefer to use a tree for the config: s3.streaming_playlists_bucket -> object_storage.streaming_playlists.bucket Use uppercase for config: S3.STREAMING_PLAYLISTS_BUCKETINFO.bucket -> OBJECT_STORAGE.STREAMING_PLAYLISTS.BUCKET (maybe BUCKET_NAME instead of BUCKET) I suggest to rename moveJobsRunning to pendingMovingJobs (or better, create a dedicated videoJobInfo table with a pendingMove & videoId columns so we could also use this table to track pending transcoding jobs) https://github.com/Chocobozzz/PeerTube/pull/4290/files#diff-3e26d41ca4bda1de8e1747af70ca2af642abcc1e9e0bfb94239ff2165acfbde5R19 uses a string instead of an integer I think we should store the origin object storage URL in fileUrl, without base_url injection. Instead, inject the base_url at "runtime" so admins can easily change this configuration without running a script to update DB URLs * Import correct function * Support multipart upload * Remove import of node 15.0 module stream/promises * Extend maximum upload job length Using the same value as for redundancy downloading seems logical * Use dynamic part size for really large uploads Also adds very small part size for local testing * Fix decreasePendingMove query * Resolve various PR comments * Move to object storage after optimize * Make upload size configurable and increase default * Prune webtorrent files that are stored in object storage * Move files after transcoding jobs * Fix federation * Add video path manager * Support move to external storage job in client * Fix live object storage tests Co-authored-by: Chocobozzz <me@florianbigard.com>
2021-08-17 01:26:20 -05:00
if (options.env) {
Object.assign(env, options.env)
}
2022-01-03 10:17:12 -06:00
const execArgv = options.nodeArgs || []
2022-01-04 02:16:43 -06:00
// FIXME: too slow :/
// execArgv.push('--enable-source-maps')
2022-01-03 10:17:12 -06:00
2021-07-16 02:47:51 -05:00
const forkOptions = {
silent: true,
env,
detached: true,
2022-01-03 10:17:12 -06:00
execArgv
2021-07-16 02:47:51 -05:00
}
2022-01-03 09:37:16 -06:00
const peertubeArgs = options.peertubeArgs || []
Add support for saving video files to object storage (#4290) * Add support for saving video files to object storage * Add support for custom url generation on s3 stored files Uses two config keys to support url generation that doesn't directly go to (compatible s3). Can be used to generate urls to any cache server or CDN. * Upload files to s3 concurrently and delete originals afterwards * Only publish after move to object storage is complete * Use base url instead of url template * Fix mistyped config field * Add rudenmentary way to download before transcode * Implement Chocobozzz suggestions https://github.com/Chocobozzz/PeerTube/pull/4290#issuecomment-891670478 The remarks in question: Try to use objectStorage prefix instead of s3 prefix for your function/variables/config names Prefer to use a tree for the config: s3.streaming_playlists_bucket -> object_storage.streaming_playlists.bucket Use uppercase for config: S3.STREAMING_PLAYLISTS_BUCKETINFO.bucket -> OBJECT_STORAGE.STREAMING_PLAYLISTS.BUCKET (maybe BUCKET_NAME instead of BUCKET) I suggest to rename moveJobsRunning to pendingMovingJobs (or better, create a dedicated videoJobInfo table with a pendingMove & videoId columns so we could also use this table to track pending transcoding jobs) https://github.com/Chocobozzz/PeerTube/pull/4290/files#diff-3e26d41ca4bda1de8e1747af70ca2af642abcc1e9e0bfb94239ff2165acfbde5R19 uses a string instead of an integer I think we should store the origin object storage URL in fileUrl, without base_url injection. Instead, inject the base_url at "runtime" so admins can easily change this configuration without running a script to update DB URLs * Import correct function * Support multipart upload * Remove import of node 15.0 module stream/promises * Extend maximum upload job length Using the same value as for redundancy downloading seems logical * Use dynamic part size for really large uploads Also adds very small part size for local testing * Fix decreasePendingMove query * Resolve various PR comments * Move to object storage after optimize * Make upload size configurable and increase default * Prune webtorrent files that are stored in object storage * Move files after transcoding jobs * Fix federation * Add video path manager * Support move to external storage job in client * Fix live object storage tests Co-authored-by: Chocobozzz <me@florianbigard.com>
2021-08-17 01:26:20 -05:00
return new Promise<void>((res, rej) => {
2021-07-16 03:20:44 -05:00
const self = this
let aggregatedLogs = ''
2021-07-16 03:20:44 -05:00
2022-01-03 09:37:16 -06:00
this.app = fork(join(root(), 'dist', 'server.js'), peertubeArgs, forkOptions)
Add support for saving video files to object storage (#4290) * Add support for saving video files to object storage * Add support for custom url generation on s3 stored files Uses two config keys to support url generation that doesn't directly go to (compatible s3). Can be used to generate urls to any cache server or CDN. * Upload files to s3 concurrently and delete originals afterwards * Only publish after move to object storage is complete * Use base url instead of url template * Fix mistyped config field * Add rudenmentary way to download before transcode * Implement Chocobozzz suggestions https://github.com/Chocobozzz/PeerTube/pull/4290#issuecomment-891670478 The remarks in question: Try to use objectStorage prefix instead of s3 prefix for your function/variables/config names Prefer to use a tree for the config: s3.streaming_playlists_bucket -> object_storage.streaming_playlists.bucket Use uppercase for config: S3.STREAMING_PLAYLISTS_BUCKETINFO.bucket -> OBJECT_STORAGE.STREAMING_PLAYLISTS.BUCKET (maybe BUCKET_NAME instead of BUCKET) I suggest to rename moveJobsRunning to pendingMovingJobs (or better, create a dedicated videoJobInfo table with a pendingMove & videoId columns so we could also use this table to track pending transcoding jobs) https://github.com/Chocobozzz/PeerTube/pull/4290/files#diff-3e26d41ca4bda1de8e1747af70ca2af642abcc1e9e0bfb94239ff2165acfbde5R19 uses a string instead of an integer I think we should store the origin object storage URL in fileUrl, without base_url injection. Instead, inject the base_url at "runtime" so admins can easily change this configuration without running a script to update DB URLs * Import correct function * Support multipart upload * Remove import of node 15.0 module stream/promises * Extend maximum upload job length Using the same value as for redundancy downloading seems logical * Use dynamic part size for really large uploads Also adds very small part size for local testing * Fix decreasePendingMove query * Resolve various PR comments * Move to object storage after optimize * Make upload size configurable and increase default * Prune webtorrent files that are stored in object storage * Move files after transcoding jobs * Fix federation * Add video path manager * Support move to external storage job in client * Fix live object storage tests Co-authored-by: Chocobozzz <me@florianbigard.com>
2021-08-17 01:26:20 -05:00
const onPeerTubeExit = () => rej(new Error('Process exited:\n' + aggregatedLogs))
2021-08-26 04:11:46 -05:00
const onParentExit = () => {
2022-11-15 08:00:19 -06:00
if (!this.app?.pid) return
2021-08-26 04:11:46 -05:00
try {
process.kill(self.app.pid)
} catch { /* empty */ }
Add support for saving video files to object storage (#4290) * Add support for saving video files to object storage * Add support for custom url generation on s3 stored files Uses two config keys to support url generation that doesn't directly go to (compatible s3). Can be used to generate urls to any cache server or CDN. * Upload files to s3 concurrently and delete originals afterwards * Only publish after move to object storage is complete * Use base url instead of url template * Fix mistyped config field * Add rudenmentary way to download before transcode * Implement Chocobozzz suggestions https://github.com/Chocobozzz/PeerTube/pull/4290#issuecomment-891670478 The remarks in question: Try to use objectStorage prefix instead of s3 prefix for your function/variables/config names Prefer to use a tree for the config: s3.streaming_playlists_bucket -> object_storage.streaming_playlists.bucket Use uppercase for config: S3.STREAMING_PLAYLISTS_BUCKETINFO.bucket -> OBJECT_STORAGE.STREAMING_PLAYLISTS.BUCKET (maybe BUCKET_NAME instead of BUCKET) I suggest to rename moveJobsRunning to pendingMovingJobs (or better, create a dedicated videoJobInfo table with a pendingMove & videoId columns so we could also use this table to track pending transcoding jobs) https://github.com/Chocobozzz/PeerTube/pull/4290/files#diff-3e26d41ca4bda1de8e1747af70ca2af642abcc1e9e0bfb94239ff2165acfbde5R19 uses a string instead of an integer I think we should store the origin object storage URL in fileUrl, without base_url injection. Instead, inject the base_url at "runtime" so admins can easily change this configuration without running a script to update DB URLs * Import correct function * Support multipart upload * Remove import of node 15.0 module stream/promises * Extend maximum upload job length Using the same value as for redundancy downloading seems logical * Use dynamic part size for really large uploads Also adds very small part size for local testing * Fix decreasePendingMove query * Resolve various PR comments * Move to object storage after optimize * Make upload size configurable and increase default * Prune webtorrent files that are stored in object storage * Move files after transcoding jobs * Fix federation * Add video path manager * Support move to external storage job in client * Fix live object storage tests Co-authored-by: Chocobozzz <me@florianbigard.com>
2021-08-17 01:26:20 -05:00
}
2021-08-26 04:11:46 -05:00
this.app.on('exit', onPeerTubeExit)
process.on('exit', onParentExit)
Add support for saving video files to object storage (#4290) * Add support for saving video files to object storage * Add support for custom url generation on s3 stored files Uses two config keys to support url generation that doesn't directly go to (compatible s3). Can be used to generate urls to any cache server or CDN. * Upload files to s3 concurrently and delete originals afterwards * Only publish after move to object storage is complete * Use base url instead of url template * Fix mistyped config field * Add rudenmentary way to download before transcode * Implement Chocobozzz suggestions https://github.com/Chocobozzz/PeerTube/pull/4290#issuecomment-891670478 The remarks in question: Try to use objectStorage prefix instead of s3 prefix for your function/variables/config names Prefer to use a tree for the config: s3.streaming_playlists_bucket -> object_storage.streaming_playlists.bucket Use uppercase for config: S3.STREAMING_PLAYLISTS_BUCKETINFO.bucket -> OBJECT_STORAGE.STREAMING_PLAYLISTS.BUCKET (maybe BUCKET_NAME instead of BUCKET) I suggest to rename moveJobsRunning to pendingMovingJobs (or better, create a dedicated videoJobInfo table with a pendingMove & videoId columns so we could also use this table to track pending transcoding jobs) https://github.com/Chocobozzz/PeerTube/pull/4290/files#diff-3e26d41ca4bda1de8e1747af70ca2af642abcc1e9e0bfb94239ff2165acfbde5R19 uses a string instead of an integer I think we should store the origin object storage URL in fileUrl, without base_url injection. Instead, inject the base_url at "runtime" so admins can easily change this configuration without running a script to update DB URLs * Import correct function * Support multipart upload * Remove import of node 15.0 module stream/promises * Extend maximum upload job length Using the same value as for redundancy downloading seems logical * Use dynamic part size for really large uploads Also adds very small part size for local testing * Fix decreasePendingMove query * Resolve various PR comments * Move to object storage after optimize * Make upload size configurable and increase default * Prune webtorrent files that are stored in object storage * Move files after transcoding jobs * Fix federation * Add video path manager * Support move to external storage job in client * Fix live object storage tests Co-authored-by: Chocobozzz <me@florianbigard.com>
2021-08-17 01:26:20 -05:00
2021-07-16 02:47:51 -05:00
this.app.stdout.on('data', function onStdout (data) {
let dontContinue = false
const log: string = data.toString()
aggregatedLogs += log
2021-07-16 02:47:51 -05:00
// Capture things if we want to
for (const key of Object.keys(regexps)) {
const regexp = regexps[key]
const matches = log.match(regexp)
2021-07-16 02:47:51 -05:00
if (matches !== null) {
2021-07-16 03:20:44 -05:00
if (key === 'client_id') self.store.client.id = matches[1]
else if (key === 'client_secret') self.store.client.secret = matches[1]
else if (key === 'user_username') self.store.user.username = matches[1]
else if (key === 'user_password') self.store.user.password = matches[1]
2021-07-16 02:47:51 -05:00
}
}
// Check if all required sentences are here
for (const key of Object.keys(serverRunString)) {
if (log.includes(key)) serverRunString[key] = true
2021-07-16 02:47:51 -05:00
if (serverRunString[key] === false) dontContinue = true
}
// If no, there is maybe one thing not already initialized (client/user credentials generation...)
if (dontContinue === true) return
if (options.hideLogs === false) {
console.log(log)
2021-07-16 02:47:51 -05:00
} else {
2021-08-26 04:11:46 -05:00
process.removeListener('exit', onParentExit)
2021-07-16 03:20:44 -05:00
self.app.stdout.removeListener('data', onStdout)
2021-08-26 04:11:46 -05:00
self.app.removeListener('exit', onPeerTubeExit)
2021-07-16 02:47:51 -05:00
}
res()
})
})
}
async kill () {
if (!this.app) return
await this.sql.cleanup()
process.kill(-this.app.pid)
this.app = null
}
private randomServer () {
const low = 10
const high = 10000
return randomInt(low, high)
}
private randomRTMP () {
const low = 1900
const high = 2100
return randomInt(low, high)
}
private async assignCustomConfigFile () {
if (this.internalServerNumber === this.serverNumber) return
const basePath = join(root(), 'config')
const tmpConfigFile = join(basePath, `test-${this.internalServerNumber}.yaml`)
await copy(join(basePath, `test-${this.serverNumber}.yaml`), tmpConfigFile)
this.customConfigFile = tmpConfigFile
}
private buildConfigOverride () {
if (!this.parallel) return {}
return {
listen: {
port: this.port
},
webserver: {
port: this.port
},
database: {
suffix: '_test' + this.internalServerNumber
},
storage: {
tmp: this.getDirectoryPath('tmp') + '/',
bin: this.getDirectoryPath('bin') + '/',
avatars: this.getDirectoryPath('avatars') + '/',
videos: this.getDirectoryPath('videos') + '/',
streaming_playlists: this.getDirectoryPath('streaming-playlists') + '/',
redundancy: this.getDirectoryPath('redundancy') + '/',
logs: this.getDirectoryPath('logs') + '/',
previews: this.getDirectoryPath('previews') + '/',
thumbnails: this.getDirectoryPath('thumbnails') + '/',
torrents: this.getDirectoryPath('torrents') + '/',
captions: this.getDirectoryPath('captions') + '/',
cache: this.getDirectoryPath('cache') + '/',
plugins: this.getDirectoryPath('plugins') + '/',
well_known: this.getDirectoryPath('well-known') + '/'
2021-07-16 02:47:51 -05:00
},
admin: {
email: `admin${this.internalServerNumber}@example.com`
},
live: {
rtmp: {
port: this.rtmpPort
}
}
}
}
private assignCommands () {
this.bulk = new BulkCommand(this)
this.cli = new CLICommand(this)
this.customPage = new CustomPagesCommand(this)
this.feed = new FeedCommand(this)
this.logs = new LogsCommand(this)
this.abuses = new AbusesCommand(this)
this.overviews = new OverviewsCommand(this)
this.search = new SearchCommand(this)
this.contactForm = new ContactFormCommand(this)
this.debug = new DebugCommand(this)
this.follows = new FollowsCommand(this)
this.jobs = new JobsCommand(this)
this.metrics = new MetricsCommand(this)
2021-07-16 02:47:51 -05:00
this.plugins = new PluginsCommand(this)
this.redundancy = new RedundancyCommand(this)
this.stats = new StatsCommand(this)
this.config = new ConfigCommand(this)
this.socketIO = new SocketIOCommand(this)
this.accounts = new AccountsCommand(this)
this.blocklist = new BlocklistCommand(this)
this.subscriptions = new SubscriptionsCommand(this)
this.live = new LiveCommand(this)
this.services = new ServicesCommand(this)
this.blacklist = new BlacklistCommand(this)
this.captions = new CaptionsCommand(this)
this.changeOwnership = new ChangeOwnershipCommand(this)
this.playlists = new PlaylistsCommand(this)
this.history = new HistoryCommand(this)
this.imports = new ImportsCommand(this)
Channel sync (#5135) * Add external channel URL for channel update / creation (#754) * Disallow synchronisation if user has no video quota (#754) * More constraints serverside (#754) * Disable sync if server configuration does not allow HTTP import (#754) * Working version synchronizing videos with a job (#754) TODO: refactoring, too much code duplication * More logs and try/catch (#754) * Fix eslint error (#754) * WIP: support synchronization time change (#754) * New frontend #754 * WIP: Create sync front (#754) * Enhance UI, sync creation form (#754) * Warning message when HTTP upload is disallowed * More consistent names (#754) * Binding Front with API (#754) * Add a /me API (#754) * Improve list UI (#754) * Implement creation and deletion routes (#754) * Lint (#754) * Lint again (#754) * WIP: UI for triggering import existing videos (#754) * Implement jobs for syncing and importing channels * Don't sync videos before sync creation + avoid concurrency issue (#754) * Cleanup (#754) * Cleanup: OpenAPI + API rework (#754) * Remove dead code (#754) * Eslint (#754) * Revert the mess with whitespaces in constants.ts (#754) * Some fixes after rebase (#754) * Several fixes after PR remarks (#754) * Front + API: Rename video-channels-sync to video-channel-syncs (#754) * Allow enabling channel sync through UI (#754) * getChannelInfo (#754) * Minor fixes: openapi + model + sql (#754) * Simplified API validators (#754) * Rename MChannelSync to MChannelSyncChannel (#754) * Add command for VideoChannelSync (#754) * Use synchronization.enabled config (#754) * Check parameters test + some fixes (#754) * Fix conflict mistake (#754) * Restrict access to video channel sync list API (#754) * Start adding unit test for synchronization (#754) * Continue testing (#754) * Tests finished + convertion of job to scheduler (#754) * Add lastSyncAt field (#754) * Fix externalRemoteUrl sort + creation date not well formatted (#754) * Small fix (#754) * Factorize addYoutubeDLImport and buildVideo (#754) * Check duplicates on channel not on users (#754) * factorize thumbnail generation (#754) * Fetch error should return status 400 (#754) * Separate video-channel-import and video-channel-sync-latest (#754) * Bump DB migration version after rebase (#754) * Prettier states in UI table (#754) * Add DefaultScope in VideoChannelSyncModel (#754) * Fix audit logs (#754) * Ensure user can upload when importing channel + minor fixes (#754) * Mark synchronization as failed on exception + typos (#754) * Change REST API for importing videos into channel (#754) * Add option for fully synchronize a chnanel (#754) * Return a whole sync object on creation to avoid tricks in Front (#754) * Various remarks (#754) * Single quotes by default (#754) * Rename synchronization to video_channel_synchronization * Add check.latest_videos_count and max_per_user options (#754) * Better channel rendering in list #754 * Allow sorting with channel name and state (#754) * Add missing tests for channel imports (#754) * Prefer using a parent job for channel sync * Styling * Client styling Co-authored-by: Chocobozzz <me@florianbigard.com>
2022-08-10 02:53:39 -05:00
this.channelSyncs = new ChannelSyncsCommand(this)
2021-07-16 02:47:51 -05:00
this.streamingPlaylists = new StreamingPlaylistsCommand(this)
this.channels = new ChannelsCommand(this)
this.comments = new CommentsCommand(this)
this.sql = new SQLCommand(this)
this.notifications = new NotificationsCommand(this)
this.servers = new ServersCommand(this)
this.login = new LoginCommand(this)
this.users = new UsersCommand(this)
this.videos = new VideosCommand(this)
Add support for saving video files to object storage (#4290) * Add support for saving video files to object storage * Add support for custom url generation on s3 stored files Uses two config keys to support url generation that doesn't directly go to (compatible s3). Can be used to generate urls to any cache server or CDN. * Upload files to s3 concurrently and delete originals afterwards * Only publish after move to object storage is complete * Use base url instead of url template * Fix mistyped config field * Add rudenmentary way to download before transcode * Implement Chocobozzz suggestions https://github.com/Chocobozzz/PeerTube/pull/4290#issuecomment-891670478 The remarks in question: Try to use objectStorage prefix instead of s3 prefix for your function/variables/config names Prefer to use a tree for the config: s3.streaming_playlists_bucket -> object_storage.streaming_playlists.bucket Use uppercase for config: S3.STREAMING_PLAYLISTS_BUCKETINFO.bucket -> OBJECT_STORAGE.STREAMING_PLAYLISTS.BUCKET (maybe BUCKET_NAME instead of BUCKET) I suggest to rename moveJobsRunning to pendingMovingJobs (or better, create a dedicated videoJobInfo table with a pendingMove & videoId columns so we could also use this table to track pending transcoding jobs) https://github.com/Chocobozzz/PeerTube/pull/4290/files#diff-3e26d41ca4bda1de8e1747af70ca2af642abcc1e9e0bfb94239ff2165acfbde5R19 uses a string instead of an integer I think we should store the origin object storage URL in fileUrl, without base_url injection. Instead, inject the base_url at "runtime" so admins can easily change this configuration without running a script to update DB URLs * Import correct function * Support multipart upload * Remove import of node 15.0 module stream/promises * Extend maximum upload job length Using the same value as for redundancy downloading seems logical * Use dynamic part size for really large uploads Also adds very small part size for local testing * Fix decreasePendingMove query * Resolve various PR comments * Move to object storage after optimize * Make upload size configurable and increase default * Prune webtorrent files that are stored in object storage * Move files after transcoding jobs * Fix federation * Add video path manager * Support move to external storage job in client * Fix live object storage tests Co-authored-by: Chocobozzz <me@florianbigard.com>
2021-08-17 01:26:20 -05:00
this.objectStorage = new ObjectStorageCommand(this)
2022-03-22 10:58:49 -05:00
this.videoStudio = new VideoStudioCommand(this)
this.videoStats = new VideoStatsCommand(this)
this.views = new ViewsCommand(this)
this.twoFactor = new TwoFactorCommand(this)
this.videoToken = new VideoTokenCommand(this)
2023-01-19 02:28:29 -06:00
this.registrations = new RegistrationsCommand(this)
2021-07-16 02:47:51 -05:00
}
}