- Ability for pod to be Completed, unlike in Statefulset - eg. if 3 pods are running and first one finishes, all 3 must be running until all 3 are done. With this setup, the first finished pod can remain in Completed state.
- Fixed shutdown order - crawler pods now correctly shutdown first before redis pods, by switching to background deletion.
- Pod priority decreases with scale: 1st instance of a new crawl can preempt 3rd or 2nd instance of another crawl
- Create priority classes upto 'max_crawl_scale, configured in values.yaml
- Improved scale change reconciliation: if increasing scale, immediately scale up. If decreasing scale,
graceful stop scaled-down instance to complete via redis 'stopone' key, wait until they exit with Completed state
before adjust status.scale / removing scaled down pods. Ensures unaccepted interrupts don't cause scaled down data to be deleted.
- Redis pod remains inactive until crawler is first active, or after no crawl pods are active for 60 seconds
- Configurable Redis storage with 'redis_storage' value, set to 3Gi by default
- CrawlJob deletion starts as soon as post-finish crawl operations are run
- Post-crawl operations get their own redis instance, since one during response is being cleaned up in finalizer
- Finalizer ignores request with incorrect state (returns 400 if reported as not finished while crawl is finished)
- Current resource usage added to status
- Profile browser: also manage single pod directly without statefulset for consistency.
- Restart pods via restartTime value: if spec.restartTime != status.restartTime, clear out pods and update status.restartTime (using OnDelete policy to avoid recreate loops in edge cases).
- Update to latest metacontroller (v4.11.0)
- Add --restartOnError flag for crawler (for browsertrix-crawler 0.11.0)
- Failed crawl logging: dd 'fail_crawl()' to be used for failing a crawl, which prints logs for default container (if enabled) as well as pod status
- tests: check other finished states to avoid stuck in infinite loop if crawl fails
- tests: disable disk utilization check, which adds unpredictability to crawl testing!
fixes#1147
---------
Co-authored-by: Tessa Walsh <tessa@bitarchivist.net>
* ingress: simplify ingress config: (fixes#1135)
- use standard Prefix pathTypes
- remove nginx-specific rewriting
- remove 'scheme', use https/http based on 'tls' setting (in ingress and configmap)
- fix signing ingress to use ingressClassName
* log only if 'log_failed_crawl_lines' value is set to number of last lines to log
from failed container
---------
Co-authored-by: Tessa Walsh <tessa@bitarchivist.net>
- set crawler cpu / memory with fixed base + incremental bumps based on number of browsers
- allow parsing k8s quantities with parse_quantity, compute in operator
- set 'crawler_cpu = crawler_cpu_base + crawler_extra_cpu_per_browser * (num_browsers - 1)'
and same for memory
* resource constraints: (fixes#895)
- for cpu, only set cpu requests
- for memory, set mem requests == mem limits
- add missing resource constraints for minio and scheduled job
- for crawler, set mem and cpu constraints per browser, scale based on browser instances per crawler
- add comments in values.yaml for crawler values being multiplied
- default values: bump crawler to 650 millicpu per browser instance just in case
cleanup: remove unused entries from main backend configmap
Operator: Modified init behavior to only load redis when at least one crawler pod available:
- waits for at least one crawler pod to be available before starting redis pod, to avoid situation where many crawler pods are in pending mode, but redis pods are still running.
- redis statefulset starts at scale of 0
- once crawler pod becomes available, redis sts is scaled to 1 (via `initRedis==true` status)
- crawl remains in 'starting' or 'waiting_capacity' state until pod becomes available without redis pod running
- set to 'running' state only after redis and at least one crawler pod is available
- if no crawler pods available after running, or, if stuck in starting for >60 seconds, switch to 'waiting_capacity' state
- when switching to 'waiting_capacity', also scale down redis to 0, wait for crawler pod to become available, only then scale up redis to 1, and get back to 'running'
other tweaks:
- add new status field 'initRedis', default to false, not displayed
- crawler pod: consider 'ContainerCreating' state as available, as container will not be blocked by resource limits
- add a resync after 3 seconds when waiting for crawler pod or redis pod to become available, configurable via 'operator_fast_resync_secs'
- set_state: if not updating state, ensure state reflects actual value in db
wabac.js will reload the replay.json on 403 with new token (will be in next version of wabac.js)
presign urls: make presign timeout configurable (in minutes), defaults to 60 mins
dockerfile: fix configuring RWP_BASE_URL
passed to crawler --userAgentSuffix and --userAgent params, respectively, using
'quote' to support spaces in user-agent.
config: re-order settings to put 'Crawler Settings' section first, followed by 'Cluster Settings'
fixes#787
* crawler args cleanup:
- move crawler args command line entirely to configmap
- add required settings like --generateWACZ and --waitOnDone to configmap to not be overridable
- values files can configure individual settings, assembled in configmap
- move disk_utilization_threshold to configmap
- add 'crawler_logging_opts' and 'crawler_extract_full_text' options to values.yaml to more easily set these options
---------
Co-authored-by: Ilya Kreymer <ikreymer@gmail.com>
* crawlconfig: fix default filename template, make configurable
- make default crawl file template configurable with 'default_crawl_filename_template' value in values.yaml
- set to '@ts-@hostsuffix.wacz' by default
- allow updating via 'crawlFilenameTemplate' in crawlconfig patch, which updates configmap
- tests: add test for custom 'default_crawl_filename_template'
* operator:
- ensures crawler pvcs are always deleted before crawl object is finalized (fixes#827)
- refactor to ensure finalizer handler always run when finalizing
- remove obsolete config entries
* Btrixjobs Operator - Phase 1 (#679)
- add metacontroller and custom crds
- add main_op entrypoint for operator
* Btrix Operator Crawl Management (#767)
* operator backend:
- run operator api in separate container but in same pod, with WEB_CONCURRENCY=1
- operator creates statefulsets and services for CrawlJob and ProfileJob
- operator: use service hook endpoint, set port in values.yaml
* crawls working with CrawlJob
- jobs start with 'crawljob-' prefix
- update status to reflect current crawl state
- set sync time to 10 seconds by default, overridable with 'operator_resync_seconds'
- mark crawl as running, failed, complete when finished
- store finished status when crawl is complete
- support updating scale, forcing rollover, stop via patching CrawlJob
- support cancel via deletion
- requires hack to content-length for patching custom resources
- auto-delete of CrawlJob via 'ttlSecondsAfterFinished'
- also delete pvcs until autodelete supported via statefulset (k8s >1.27)
- ensure filesAdded always set correctly, keep counter in redis, add to status display
- optimization: attempt to reduce automerging, by reusing volumeClaimTemplates from existing children, as these may have additional props added
- add add_crawl_errors_to_db() for storing crawl errors from redis '<crawl>:e' key to mongodb when crawl is finished/failed/canceled
- add .status.size to display human-readable crawl size, if available (from webrecorder/browsertrix-crawler#291)
- support new page size, >0.9.0 and old page size key (changed in webrecorder/browsertrix-crawler#284)
* support for scheduled jobs!
- add main_scheduled_job entrypoint to run scheduled jobs
- add crawl_cron_job.yaml template for declaring CronJob
- CronJobs moved to default namespace
* operator manages ProfileJobs:
- jobs start with 'profilejob-'
- update expiry time by updating ProfileJob object 'expireTime' while profile is active
* refactor/cleanup:
- remove k8s package
- merge k8sman and basecrawlmanager into crawlmanager
- move templates, k8sapi, utils into root package
- delete all *_job.py files
- remove dt_now, ts_now from crawls, now in utils
- all db operations happen in crawl/crawlconfig/org files
- move shared crawl/crawlconfig/org functions that use the db to be importable directly,
including get_crawl_config, add_new_crawl, inc_crawl_stats
* role binding: more secure setup, don't allow crawler namespace any k8s permissions
- move cronjobs to be created in default namespace
- grant default namespace access to create cronjobs in default namespace
- remove role binding from crawler namespace
* additional tweaks to templates:
- templates: split crawler and redis statefulset into separate yaml file (in case need to load one or other separately)
* stats / redis optimization:
- don't update stats in mongodb on every operator sync, only when crawl is finished
- for api access, read stats directly from redis to get up-to-date stats
- move get_page_stats() to utils, add get_redis_url() to k8sapi to unify access
* Add migration for operator changes
- Update configmap for crawl configs with scale > 1 or
crawlTimeout > 0 and schedule exists to recreate CronJobs
- add option to rerun last migration, enabled via env var and by running helm with --set=rerun_last_migration=1
* subcharts: move crawljob and profilejob crds to separate subchart, as this seems best way to guarantee proper install order with + update on upgrade with helm, add built btrix-crds-0.1.0.tgz subchart
- metacontroller: use release from ghcr, add metacontroller-helm-v4.10.1.tgz subchart
* backend api fixes
- ensure changing scale of crawl also updates it in the db
- crawlconfigs: add 'currCrawlSize' and 'lastCrawlSize' to crawlconfig api
---------
Co-authored-by: D. Lee <leepro@gmail.com>
Co-authored-by: Tessa Walsh <tessa@bitarchivist.net>
* config: add 'pageLoadTimeout' and 'pageExtraDelay' options to backend config
- add 'default_page_load_timeout_seconds' to values.yaml, defaulting to 120, for pageLoadTimeout
- add 'defaultPageLoadTimeSeconds ' to /api/settings, update tests for /api/settings
addresses issue in #636
* more page limit: update to #717, instead of setting --limit in each crawlconfig,
apply override --maxPageLimit setting, implemented in crawler, to override individually configured page limit
* update tests, no longer returning 'crawl_page_limit_exceeds_allowed'
* backend: max pages per crawl limit, part of fix for #716:
- set 'max_pages_crawl_limit' in values.yaml, default to 100,000
- if set/non-0, automatically set limit if none provided
- if set/non-0, return 400 if adding config with limit exceeding max limit
- return limit as 'maxPagesPerCrawl' in /api/settings
- api: /all/crawls - add runningOnly=0 to show all crawls, default to 1/true (for more reliable testing)
tests: add test for 'max_pages_per_crawl' setting
- ensure 'limit' can not be set higher than max_pages_per_crawl
- ensure pages crawled is at the limit
- set test limit to max 2 pages
- add settings test
- check for pages.jsonl and extraPages.jsonl when crawling 2 pages
* chart crawl args cleanup:
- move configurable settings out of 'crawler_args'
- add 'crawler_session_size_limit_bytes' and 'crawler_session_time_limit_seconds' for --timeLimit and --sizeLimit option for crawler
- remove hard-coded 'timeout' to allow configuring via crawl config
- set liveness check port from existing config value
- add comments that requests hd must be at least double the size limit
- defaults: set crawler_requests_hd to 22GB, default crawl session size limit to 10GB
- partial fix to #321, don't hard-code behavior limit into crawler args
- allow setting number of crawler browser instances via 'crawler_browser_instances' to avoid having to override the full crawler args
* Make invites expire after configurable window
The value can be set in EXPIRE_AFTER_SECONDS env var and via
helm chart values, and defaults to 7 days.
* Create nightly test CI and add invite expiration test to it
* Update 404 error message for missing or expired invite
---------
Co-authored-by: sua yoo <sua@suayoo.com>
* rename config values
- api -> backend
- nginx -> frontend
* job pods:
- set job_pull_policy from api_pull_policy (same as backend image)
- default to Always, but can be overridden for local deployment (same as backend image)
typo fix: CRAWL_NAMESPACE -> CRAWLER_NAMESPACE (part of #491)
ansible: set default label to :latest instead of :dev for
- Add default switch to Archive (org) model
- Set default org name via values.yaml
- Add check to ensure only one org with default org name exists
- Stop creating new orgs for new users
- Add new API endpoints for creating and renaming orgs (part of #457)
- Make Archive.name unique via index
- Wait for db connection on init, log if waiting
- Make archive-less invites invite user to default org with Owner role
- Rename default org from chart value if changed
- Don't create new org for invited users
* profile browser vnc support + fixes:
- switch profile browser rendering to use VNC
- frontend: add @novnc/novnc as dependency, create separate bundle novnc.js to load into vnc browser (to avoid loading from each container)
- frontend: update proxy paths to proxy websocket, index page to crawler
- frontend: allow browser profiles in all browsers, remove browser compatibility check
- frontend: update webpack dev config, apply prettier
- frontend: node version fix
- backend: get vncpassword, build new URL for proxying to crawler iframe
- backend: fix profile / crawl job pull policy from 'Always' -> 'Never', should use existing image for job
- backend: fix kill signal to use bash -c to work with latest backend image
- backend/chart: add 'profile_browser_timeout_seconds' to chart values to control how long profile browser to remain when idle (default to 60)
- backend: remove utils.py, now using secret.token_hex() for random suffix
Co-authored-by: sua yoo <sua@suayoo.com>
- ingress: fix proxying /data to minio, use another ingress which proxies correct host to ensure presigned urls work
- presigning: determine if signing endpoint url (minio) or access endpoint (cloud bucket) based on if access endpoint is provided, set bool on storage object
- chart: fix indent on incorrect storageClassName configs
- ingress: make 'ingress_class' configurable (set to 'public' for microk8s, default to 'nginx')
- minio: use older minio image which supports legacy fs based setup (for now)
- nginx service: add 'nginx_service_use_node_port' config setting: if true, will use NodePort for frontend,
other will use default (ClusterIP) and only for the frontend / nginx
- chart: remove changing service type for other services
- prefer deploy crawler, redis and job to same zone
- prefer deploying crawler and job together via crawler node type, redis via redis node type (all optional)
- use python-on-whale to use docker cli api directly, creating docker stack for each crawl or profile browser
- configure storages via storages.yaml secret
- add crawl_job, profile_job, splitting into base and k8s/swarm implementations
- split manager into base crawlmanager and k8s/swarm implementations
- swarm: load initial scale from db to avoid modifying fixed configs, in k8s, load from configmap
- swarm: support scheduled jobs via swarm-cronjob service
- remove docker dependencies (aiodocker, apscheduler, scheduling)
- swarm: when using local minio, expose via /data/ route in nginx via extra include (in k8s, include dir is empty and routing handled via ingress)
- k8s: cleanup minio chart: move init containers to minio.yaml
- swarm: stateful set implementation to be consistent with k8s scaling:
- don't use service replicas,
- create a unique service with '-N' appended and allocate unique volume for each replica
- allows crawl containers to be restarted w/o losing data
- add volume pruning background service, as volumes can be deleted only after service shuts down fully
- watch: fully simplify routing, route via replica index instead of ip for both k8s and swarm
- rename network btrix-cloud-net -> btrix-net to avoid conflict with compose network
- use statefulsets instead of deployments for mongo, redis, signer
- use k8s job + statefulset for running crawls
- use separate statefulset for crawl (scaled) and single-replica redis stateful set
- move crawl job update login to crawl_updater
- remove shared redis chart
package refactor:
- move to shared code to 'btrixcloud'
- move k8s to 'btrixcloud.k8s'
- move docker to 'btrixcloud.docker'
* backend: k8s:
- support crawls with multiple wacz files, don't assume crawl complete after first wacz uploaded
- if crawl is running and has wacz file, still show as running
- k8s: allow configuring node selector for main pods (eg. nodeType=main) and for crawlers (eg. nodeType=crawling)
- profiles: support uploading to alternate storage specified via 'shared_profile_storage' value is set
- misc fixes for profiles
* backend: ensure docker run_profile api matches k8s
k8s chart: don't delete pvc and pv in helm chart
* dependency: bump authsign to 0.4.0
docker: disable public redis port
* profiles: fix path, profile browser return value
* fix typo in presigned url cacheing
- add liveness probe for crawls, configurable via 'crawler_liveness_port'
- add User system:anonymous permissions
- treat jobs that have exceeded total as 'partial_complete' (experimental)
* backend: makes mongo auth configurable!
use mongo_auth secret in k8s and set env vars in docker
fixes#177
* docker: update config.sample.env: use ws screencast by default, add NO_DELETE_ON_FAIL option, extend default login lifetime
- set WEB_CONCURRENCY env var to configure number of backend api workers for both docker and k8s
- set via 'backend_workers' in values.yaml
- also add 'rwp_base_url' to values.yaml
- update containers to use public webrecorder/browsertrix-backend and webrecorder/browsertrix-frontend containers
- make liveness, readiness and startup health checks more tolerant
- set resource mem and cpu requests/limits for all used services (not minio for now)
- add readiness proble to redis, mongo
- adjust crawler limits, set via configmap
- add 'emptyDir' volume for crawl directory (to allow any pod restarts to have access to the data)
- rename minio and redis volumes to avoid any confusion
- add pod termination grace-period (default to 600 secs)
* backend fixes: fix graceful stop + stats
- use redis to track stopping state, to be overwritten when finished
- also include stats in completed crawls
- docker: use short container id for crawl id
- graceful stop returns 'stopping_gracefully' instead of 'stopped_gracefully'
- don't set stopping state when complete!
- beginning files support: resolve absolute urls for crawl detail (not pre-signing yet)
* backend:
- refactor invite system, move to separate InviteOps object, used by archives and user
- supporting three invite use cases:
1) superuser invites any user not registered, not added to any archive
2) archive admin invites any user not registered, add to one of their archives
3) archive admin invites existing registered user, add to one of their archives
- support superadmin invite via /users/invite (fixes#37)
- superadmin invite has no archive set and does not add user to archive
- don't send verification email when accepting from invite, fixes#50
- use different email template / accept url for existing user invite, eg, `/invite/accept/`
- fix default token value in chart
* backend:
- add /api/settings endpoint for misc system-wide settings
- setting 'registrationEnabled' if open registration should be enabled, set via REGISTRATION_ENABLED=1 env var
- setting 'jwtTokenLifetimeMinutes' returns the jwt token expiry in seconds, configured in minutes via JWT_TOKEN_LIFETIME_MINUTES env var (default: 60)
* support running backend + frontend together on k8s
* split nginx container into separate frontend service, which uses nignx-base image and the static frontend files
* add nginx-based frontend image to docker-compose build (for building only, docker-based combined deployment not yet supported)
* backend:
- fix paths for email templates
- chart: support '--set backend_only=1' and '--set frontend_only=1' to only force deploy one or the other
- run backend from root /api in uvicorn
* k8s: support email configuration
support sending reset password email
fix for #32
* fastapi users: update to latest (8.1.2)
send verification email upon registration
* update to latest fastapi-users(8.1.2), refactor to use UserManager class
ensure verification e-mail sent upon registration, w/o requiring separate apicall
fixes#32
* add email options to default chart/values.yaml
* separate usermanager init from fastapi users init, fix for sending invite emails
* misc backend fixes:
- fix running w/o local minio
- ensure crawler image pull policy is configurable, loaded via chart value
- use digitalocean repo for main backend image (for now)
- add bucket_name to config only if using default bucket
* enable all behaviors, support 'access_endpoint_url' for default storages
* debugging: add 'no_delete_jobs' setting for k8s and docker to disable deletion of completed jobs
support screencasting to dynamically created service via nginx (k8s only thus far)
add crawl /watch endpoint to enable watching, creates service if doesn't exist
add crawl /running endpoint to check if crawl is running
nginx auth check in place, but not yet enabled
add k8s nginx.conf
add missing chart files
file reorg: move docker config to configs/
k8s: add readiness check for nginx and api containers for smoother reloading
ensure service deleted along with job
todo: update dockerman with screencast support
move mongo into separate optional deployment along with minio
support for configuring storages
support for deleting crawls, associated config and secrets
- working apis for adding crawls, removing crawls in mongo, mapped to k8s cronjobs
- more complete crawl spec
- option to start on-demand job from cronjobs
- optional minio in separate deployment/service
