GlympseVideo/Glympse_Docs
1
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity

new pages
Some checks failed
Deploy Hugo site to Pages / build (push) Failing after 39s
Details
Deploy Hugo site to Pages / deploy (push) Has been skipped
Details

Browse Source
This commit is contained in:
Joshua Kirkcaldy
2024-11-26 17:28:08 +00:00
parent 43f9a55f48
commit e7b0cbdbf6
34 changed files with 593 additions and 65 deletions
Download Patch File Download Diff File Expand all files Collapse all files

3
content/docs/Installation/database.md
View File

@@ -73,9 +73,8 @@ services:
networks:
glympse_net: null
networks:
default:
glympse_net:
external: true
name: glympse_net
```
There are two more files that you need to create and save before running the compose file.

150
content/docs/Installation/install_glympse.md
View File

@@ -12,14 +12,49 @@ UWSGI and NGINX are required for the webui service
The following compose file will enable all services in a single container. This will work but it is possible that the transcoding and transcribing tasks can slow down the webui if you're running on a slower machine.
If you are going to be running the transcoder or the transcription engine, you will need an Nvidia GPU. To find the device ID run `nvidia_smi` on the system you will be installing the container.
This will give you an output that looks like this:
```
+-----------------------------------------------------------------------------------------+
| NVIDIA-SMI 565.57.02 Driver Version: 566.03 CUDA Version: 12.7 |
|-----------------------------------------+------------------------+----------------------+
| GPU Name Persistence-M | Bus-Id Disp.A | Volatile Uncorr. ECC |
| Fan Temp Perf Pwr:Usage/Cap | Memory-Usage | GPU-Util Compute M. |
| | | MIG M. |
|=========================================+========================+======================|
| 0 NVIDIA GeForce GTX 1080 On | 00000000:01:00.0 On | N/A |
| 27% 38C P8 8W / 180W | 893MiB / 8192MiB | 1% Default |
| | | N/A |
+-----------------------------------------+------------------------+----------------------+
+-----------------------------------------------------------------------------------------+
| Processes: |
| GPU GI CI PID Type Process name GPU Memory |
| ID ID Usage |
|=========================================================================================|
| 0 N/A N/A 26 G /Xwayland N/A |
+-----------------------------------------------------------------------------------------+
```
The device id is in the third row, just above the memory:
```
+========================+
| 00000000:01:00.0 On |
| 893MiB / 8192MiB |
| |
+------------------------+
```
In this case the device id is: `00000000:01:00.0`
Add this to your compose file. A file that runs all services with a gpu looks like:
```yaml {filename="compose.yml"}
services:
Glympse:
Glympse_net:
image: git.themainframe.co.uk/josh/glympse
container_name: Glympse
restart: unless-stopped
ports:
- 80:80
- 8080:80
environment:
TZ: Europe/London
ENABLE_UWSGI: yes
@@ -29,11 +64,45 @@ services:
ENABLE_MIGRATIONS: yes
ENABLE_TRANSCRIBER: yes
ENABLE_TRANSCODER: yes
DJANGO_SETTINGS_MODULE: Glympse.settings.production
SECRET_KEY: "super_secret_key"
SECRET_ADMIN_URL: random_string
CSRF_TRUSTED_ORIGINS: http://127.0.0.1
ALLOWED_HOSTS: 127.0.0.1
ADMINS: admin@glympsevideo.com
DOMAIN: glympse.glympsevideo.com
SQL_ENGINE: django.db.backends.mysql
SQL_DATABASE: glympse
SQL_USER: glympse
SQL_PASSWORD: insecure_password
SQL_HOST: mysql
SQL_PORT: 3306
EMAIL_HOST: mail.themainframe.co.uk
EMAIL_PORT: 587
EMAIL_USE_TLS: true
EMAIL_HOST_USER: noreply@glympsevideo.com
EMAIL_HOST_PASSWORD: insecure_password
DEFAULT_FROM_EMAIL: Glympse<glympse@glympsevideo.com>
ALLOWED_EMAIL_DOMAINS: glympsevideo.com
REDIS_HOST: redis://redis:6379
RABBITMQ_URL: amqp://glympse:insecure_password@rabbitmq:5672/glympse
ADMIN_USERNAME: hub
ADMIN_EMAIL: admin@glympsevideo.com
ADMIN_PASSWORD: insecure_password
LOG_LEVEL: DEBUG
TIMEZONE: Europe/London
MICROSOFT_AUTH_CLIENT_ID: None
MICROSOFT_AUTH_CLIENT_SECRET: None
MICROSOFT_AUTH_TENANT_ID: None
MEMCACHED: memcached:11211
CUDA_DEVICE: cuda:0
DEBUG: "True"
REMOTE_WORKER: "False"
REMOTE_PRODUCTION: None
volumes:
- /opt/glympse/logs:/Glympse/logs
- /opt/glympse/.env:/Glympse/config/.env
- <Glympse media location>:/media
- <rushes location>:/rushes:ro
- /media:/media
- /rushes:/rushes:ro
deploy:
resources:
reservations:
@@ -42,56 +111,43 @@ services:
device_ids:
- 00000000:01:00.0
capabilities:
- gpu
- gpu
networks:
glympse_net: null
memcached:
image: memcached:latest
container_name: memcached
entrypoint: memcached -m 256
restart: unless-stopped
restart: unless-stopped
networks:
glympse_net: null
networks:
default:
glympse_net:
external: true
name: glympse_net
```
If you are install Glympse on a second system or a system other than where the raw camera files are stored you can use the following to mount the volumes as smb shares:
```yaml
volumes:
rushes:
driver: local
driver_opts:
type: cifs
o: username=<sbm_username>,password=<smb_password>,ro,domain=localhost
device: \\smb\share\path
```
The following env file is required. You will need to change some of the options to what you specified in the database section of the install.
Alternatively you can use a nfs share:
```yaml
volumes:
rushes:
driver_opts:
type: nfs
o: "addr=<nfs-server-ip>,nolock,soft,rw"
device: ":/full/share/path"
```
Every line must be in the env file for every container you run regardless of what services are enabled.
You will need to make sure that if you are mounting the raw rushes volume that you mount this as `ro` or read only so Glympse can not make any changes to the raw rushes.
```env {filename=".env"}
DJANGO_SETTINGS_MODULE="Glympse.settings.production"
SECRET_KEY="<generate a long random string>"
SECRET_ADMIN_URL=<Generate a short random string>
CSRF_TRUSTED_ORIGINS=https://"<Glympse domain>"
ALLOWED_HOSTS="<Glympse Domain>"
ADMINS='<admin email address>'
DOMAIN='<Glympse domain>'
MAX_FIELDS=5000
SQL_ENGINE="django.db.backends.mysql"
SQL_DATABASE="glympse"
SQL_USER="glympse"
SQL_PASSWORD="<your glympse mysql password>"
SQL_HOST="mysql"
SQL_PORT="3306"
EMAIL_HOST="<mail server>"
EMAIL_PORT="<mail port>"
EMAIL_USE_TLS="<mail tls>"
EMAIL_HOST_USER="<email username>"
EMAIL_HOST_PASSWORD="<email password>"
DEFAULT_FROM_EMAIL="Glympse<glympse@example.com>"
ALLOWED_EMAIL_DOMAINS="<allowed domains for user sign up>"
REDIS_HOST="redis://redis:6379"
RABBITMQ_URL="amqp://glympse:<rabbitmq password>@rabbitmq:5672/glympse"
ADMIN_USERNAME="<admin username>"
ADMIN_EMAIL="<admin email address>"
ADMIN_PASSWORD="<default admin password>"
LOG_LEVEL="INFO"
TIMEZONE="Europe/London"
MICROSOFT_AUTH_CLIENT_ID='<client ID for SSO>'
MICROSOFT_AUTH_CLIENT_SECRET='<client secret for SSO>'
MICROSOFT_AUTH_TENANT_ID='<tenant ID for SSO>'
MEMCACHED='memcached:11211'
CUDA_DEVICE='cuda:0'
DEBUG=False
```
If you are mounting the glympse mediafiles or logs, you will need to mount this as `rw` or read write so Glympse can create the files.

3
content/docs/Installation/services.md
View File

@@ -8,6 +8,9 @@ next: docs/install_glympse
Glympse is split into multiple services so you can split the workload between multiple machines.
The available services are as follows:
* Celery Beat\
You should only have one of these running across all your containers. This is responsible for managing the scheduled tasks. Without this, the scheduled tasks will not run. For ease, I usually run this in the same container as the webui.
* WebUI\
This is the webui front end. The reverse proxy needs to point to the machine that is running this service. It is possible to run more than one frontend for load balancing purposes.