Comments (9)
CPU usage comes down to indicator calculation. I'm running 15-21 bots on a single server and use about 30-40GB RAM (5m timeframe plus informatives) and need to use 3 binance-proxy services to make sure I don't hit rate limits.
from freqtrade.
You can split the config into several parts, so you only have to change certain part, and use same files for the rest
from freqtrade.
How does that ensure 1 api call to the exchange or does it when the each strategy has its own terminal?
from freqtrade.
If you want only 1 API call, use proxy. Google "freqtrade binance proxy" or "freqtrade kucoin proxy"
from freqtrade.
Please get used to filling out the issue template. Those who frequently omit the template may have their issues closed without comment.
We're asking some very basic questions - by deleting the issue template, you're explicitly choosing not to answer our questions.
Filling out the template helps reduce our workload, and we greatly appreciate that.
While it may seem identical - calls are usually not.
Timing will be different - and calls deviate a lot once you're in a position with one bot, but not with the other.
One IP can usually handle ~3 bots (on binance, at least) - after that, you'll need to scale horizontally, using multiple servers.
Usually however, the limit is also the resources available on the actual system - not just api calls - so scaling horizontally (more servers) usually makes sense after a few bots - unless you attempt to run on a pretty big server / system).
Whilest running with a proxy is something common - all of the ones i looked at are unmaintained - and we will be unable to support deployments that do use such proxies.
from freqtrade.
Noted, Idk, but the limits of running multiple strategies on a single machine are quickly reached. I am dry run testing 10 strategies and I am using about 16GB ram and CPU is at 100%.
If any efficiencies are available in the next versions they are welcome.
Like for instance, a single call to the exchange.
from freqtrade.
Call to exchange don't contribute a lot to the CPU usage. Your strategies are too "heavy". Make them simpler if you can
from freqtrade.
Idk what the solution is here. Compile the strategies? Shared candlestick data? If the performance of dry run testing could improve that would allow for running more strategies.
pmap -X pid lists heap and a strategy is running checks on the heap every 0.1ms
Flat files? for dry run mode.
from freqtrade.
Shared candlestick data
- that's what an exchange proxy like binance-proxy does. You still have to do the indicator calculation per bot instance though, so again, you're not really saving many CPU cycles. Compiling strategies will make zero difference to this. There's no real "issue" with the performance of dry runs. The more complex your strategy is, the higher single core clock speed you'll need.
from freqtrade.
Related Issues (20)
- modify the method to guarantee the execution of model training (freqai_interface - ensure_data_exists) HOT 2
- Using tuple in CategoricalParameter HOT 6
- Access to metadta in bot_start callback HOT 2
- Custom_entry_price works not as expected? HOT 21
- adjust_trade_position does not exit Trade HOT 6
- Backtesting: KeyError: 'ME' HOT 5
- Trade object: Remaining amount HOT 4
- Hyperopt finished loading data, the epoch phase stayed at 0% and exited the run HOT 4
- AI Code optimization "proof-reading"? HOT 4
- How to enable sandbox mode on OKX exchange HOT 2
- Hyperopt loss going wrong way ? HOT 8
- Is it good practice to store objects in main dataframe? HOT 6
- Strategy backtesting without error alerts, auto-stop HOT 3
- Impossible to backtest on Bybit futures HOT 11
- Is it possible to plot bool indicators below or above candles & plot shapes? HOT 5
- Different results with same parameters HOT 4
- TEMA200 to long? HOT 10
- Custom Telegram message from Strategy HOT 1
- ccxt >= 4.3.21 is broken HOT 1
- FreqAI: model not retraining itself. HOT 3
Recommend Projects
-
React
A declarative, efficient, and flexible JavaScript library for building user interfaces.
-
Vue.js
🖖 Vue.js is a progressive, incrementally-adoptable JavaScript framework for building UI on the web.
-
Typescript
TypeScript is a superset of JavaScript that compiles to clean JavaScript output.
-
TensorFlow
An Open Source Machine Learning Framework for Everyone
-
Django
The Web framework for perfectionists with deadlines.
-
Laravel
A PHP framework for web artisans
-
D3
Bring data to life with SVG, Canvas and HTML. 📊📈🎉
-
Recommend Topics
-
javascript
JavaScript (JS) is a lightweight interpreted programming language with first-class functions.
-
web
Some thing interesting about web. New door for the world.
-
server
A server is a program made to process requests and deliver data to clients.
-
Machine learning
Machine learning is a way of modeling and interpreting data that allows a piece of software to respond intelligently.
-
Visualization
Some thing interesting about visualization, use data art
-
Game
Some thing interesting about game, make everyone happy.
Recommend Org
-
Facebook
We are working to build community through open source technology. NB: members must have two-factor auth.
-
Microsoft
Open source projects and samples from Microsoft.
-
Google
Google ❤️ Open Source for everyone.
-
Alibaba
Alibaba Open Source for everyone
-
D3
Data-Driven Documents codes.
-
Tencent
China tencent open source team.
from freqtrade.