WIP: Run Knight Crawler
This commit is contained in:
@@ -18,4 +18,46 @@ Rename the <path>.env.example</path> file to be <path>.env</path>
|
|||||||
└── docker-compose.yaml
|
└── docker-compose.yaml
|
||||||
```
|
```
|
||||||
|
|
||||||
Before we start the services, we need to
|
## Configuration
|
||||||
|
|
||||||
|
Before we start the services, we need to change a few things in the <path>.env</path> file.
|
||||||
|
|
||||||
|
> If you are using an external database, configure it in the <path>.env</path> file. Don't forget to disable the ones
|
||||||
|
> included in the <path>docker-compose.yaml</path>.
|
||||||
|
|
||||||
|
### Your time zone.
|
||||||
|
|
||||||
|
```Bash
|
||||||
|
TZ=London/Europe
|
||||||
|
```
|
||||||
|
|
||||||
|
> A list of time zones can be found on [Wikipedia](https://en.wikipedia.org/wiki/List_of_tz_database_time_zones)
|
||||||
|
|
||||||
|
### Consumers
|
||||||
|
|
||||||
|
```Bash
|
||||||
|
JOB_CONCURRENCY=5
|
||||||
|
...
|
||||||
|
MAX_CONNECTIONS_PER_TORRENT=10
|
||||||
|
...
|
||||||
|
CONSUMER_REPLICAS=3
|
||||||
|
```
|
||||||
|
|
||||||
|
These are totally subjective to your machine and network capacity. The above default is pretty minimal and will work on most machines.
|
||||||
|
|
||||||
|
`JOB_CONCURRENCY` is how many films and tv shows the consumers should process at once. As this affects every consumer this will likely cause exponential
|
||||||
|
strain on your system. It's probably best to leave this at 5, but you can try experimenting with it if you wish.
|
||||||
|
|
||||||
|
`MAX_CONNECTIONS_PER_TORRENT` is how many peers the consumer will attempt to connect to when it is trying to collect metadata.
|
||||||
|
Increasing this value can speed up processing, but you will eventually reach a point where more connections are being made than
|
||||||
|
your router can handle. This will then cause a cascading fail where your internet stops working. If you are going to increase this value
|
||||||
|
then try increasing it by 10 at a time.
|
||||||
|
|
||||||
|
> Increasing this value increases the max connections for every parallel job for every consumer. For example
|
||||||
|
> with the default values above this means that Knight Crawler will be on average making `(5 x 3) x 10 = 150` connections at any one time.
|
||||||
|
>
|
||||||
|
{style="warning"}
|
||||||
|
|
||||||
|
`CONSUMER_REPLICAS` is how many consumers should be started. This is the ultimate decider in how fast you will be able to
|
||||||
|
add films and tv shows to your database. However, this is also going to be the most intensive service you will run.
|
||||||
|
The default of 3 is a reasonable starting amount. It will work on almost every system.
|
||||||
Reference in New Issue
Block a user