How many CPUs are needed for smooth operation in quote mode?
I run the active WL, which sends the signals to the exchange, in a VPS. When I start 2 quotes (Binance and Kraken with 30 stop orders each) the CPUs are loaded over 90%. Keys or mouse inputs are only used after 10-20 seconds, if at all.
Then I increased the number of CPUs to 4. That did not help either.
If I double again, I would have to pay over $250 per month.
Is there a solution within WL?
If not I am thinking about booking WL a second time and separating shares and cryptos. Then I wouldn't have to keep changing the settings back and forth in "Backtest".
Does anyone have any ideas?
I run the active WL, which sends the signals to the exchange, in a VPS. When I start 2 quotes (Binance and Kraken with 30 stop orders each) the CPUs are loaded over 90%. Keys or mouse inputs are only used after 10-20 seconds, if at all.
Then I increased the number of CPUs to 4. That did not help either.
If I double again, I would have to pay over $250 per month.
Is there a solution within WL?
If not I am thinking about booking WL a second time and separating shares and cryptos. Then I wouldn't have to keep changing the settings back and forth in "Backtest".
Does anyone have any ideas?
Rename
Sounds like we need to take a look at it and get back to you. Please post the lists of symbols you're using.
Could you please copy the symbols from the DataSet and paste them in post so that I can monitor the same DataSets. Thanks and sorry for the bother!
Kraken:
ADA.EUR ALGO.EUR APE.EUR ATOM.EUR AVAX.EUR AXS.EUR BCH.EUR BIT.EUR BTC.EUR DOGE.EUR DOT.EUR ETH.EUR GALA.EUR GMT.EUR ICP.EUR LINK.EUR LTC.EUR LUNA.EUR MATIC.EUR NEAR.EUR QNT.EUR SAND.EUR SHIB.EUR SOL.EUR TRX.EUR UNI.EUR XLM.EUR XMR.EUR XRP.EUR
ADA.EUR ALGO.EUR APE.EUR ATOM.EUR AVAX.EUR AXS.EUR BCH.EUR BIT.EUR BTC.EUR DOGE.EUR DOT.EUR ETH.EUR GALA.EUR GMT.EUR ICP.EUR LINK.EUR LTC.EUR LUNA.EUR MATIC.EUR NEAR.EUR QNT.EUR SAND.EUR SHIB.EUR SOL.EUR TRX.EUR UNI.EUR XLM.EUR XMR.EUR XRP.EUR
Binance:
ADA.BTC ALGO.BTC APE.BTC ATOM.BTC AVAX.BTC BCH.BTC BNB.BTC DOGE.BTC DOT.BTC ETC.BTC ETH.BTC FIL.BTC FLOW.BTC FTT.BTC HBAR.BTC ICP.BTC LINK.BTC LTC.BTC MANA.BTC MATIC.BTC NEAR.BTC SAND.BTC SOL.BTC TRX.BTC UNI.BTC VET.BTC XLM.BTC XMR.BTC XRP.BTC XTZ.BTC
ADA.BTC ALGO.BTC APE.BTC ATOM.BTC AVAX.BTC BCH.BTC BNB.BTC DOGE.BTC DOT.BTC ETC.BTC ETH.BTC FIL.BTC FLOW.BTC FTT.BTC HBAR.BTC ICP.BTC LINK.BTC LTC.BTC MANA.BTC MATIC.BTC NEAR.BTC SAND.BTC SOL.BTC TRX.BTC UNI.BTC VET.BTC XLM.BTC XMR.BTC XRP.BTC XTZ.BTC
Binance mapping:
[[\.\/\-]=]
DOT.BTC=DOTBTC
[[\.\/\-]=]
DOT.BTC=DOTBTC
Kraken mapping:
[[\.\/\-]=]
XRP.EUR=XXRPZEUR
ETH.EUR=XETHZEUR
BTC.EUR=XXBTZEUR
XBTZEUR=XXBTZEUR
LTC.EUR=XLTCZEUR
DOGE.EUR=XDGEUR
ETC.EUR=XETCZEUR
XLM.EUR=XXLMZEUR
XMR.EUR=XXMRZEUR
XRPEUR=XXRPZEUR
ETHEUR=XETHZEUR
BTCEUR=XXBTZEUR
XBTZEUR=XXBTZEUR
LTCEUR=XLTCZEUR
DOGEEUR=XDGEUR
ETCEUR=XETCZEUR
XLMEUR=XXLMZEUR
XMREUR=XXMRZEUR
[[\.\/\-]=]
XRP.EUR=XXRPZEUR
ETH.EUR=XETHZEUR
BTC.EUR=XXBTZEUR
XBTZEUR=XXBTZEUR
LTC.EUR=XLTCZEUR
DOGE.EUR=XDGEUR
ETC.EUR=XETCZEUR
XLM.EUR=XXLMZEUR
XMR.EUR=XXMRZEUR
XRPEUR=XXRPZEUR
ETHEUR=XETHZEUR
BTCEUR=XXBTZEUR
XBTZEUR=XXBTZEUR
LTCEUR=XLTCZEUR
DOGEEUR=XDGEUR
ETCEUR=XETCZEUR
XLMEUR=XXLMZEUR
XMREUR=XXMRZEUR
I need the list of symbols you were showing in the Quote windows. Even so, using the lists that you gave me with the mappings, and Quotes open for Binance and Kraken my cpu utilization moves between about 10% and 25% peaks.
Mine isn't a VPS, but it's also not the latest and greatest processor either - Intel(R)Core(TM)iy-4770S CPU @ 3.10GHz. What's the difference?
I'll try again with the symbol lists you were using in your images, but you have to give them to me. Your DataSets are not the ones above with the period separators you supplied. Your trading lists are different.
Mine isn't a VPS, but it's also not the latest and greatest processor either - Intel(R)Core(TM)iy-4770S CPU @ 3.10GHz. What's the difference?
I'll try again with the symbol lists you were using in your images, but you have to give them to me. Your DataSets are not the ones above with the period separators you supplied. Your trading lists are different.
Symbols:
The list in the course window is generated by the mapping. The base data is from Cryptocompare and is mapped into the symbols of Kraken and Binance respectively.
Does the mapping possibly consume performance?
Machine:
Maybe it is due to the number of cores. More cores share the work better.
I think the performance of VPS is always a bit lower, although I have a dedicated VPS that is already a bit better.
The list in the course window is generated by the mapping. The base data is from Cryptocompare and is mapped into the symbols of Kraken and Binance respectively.
Does the mapping possibly consume performance?
Machine:
Maybe it is due to the number of cores. More cores share the work better.
I think the performance of VPS is always a bit lower, although I have a dedicated VPS that is already a bit better.
The problem still exists, does anyone have any ideas?
Does anyone else even see the problem? Lukas, do you even see the problem on your non-VPS machine? Maybe the VPS needs more "oomph" - excuse my highly technical language.
Lukas, could you tell us how the VPS is configured (CPU, RAM)?
QUOTE:
Does anyone else even see the problem? Lukas, do you even see the problem on your non-VPS machine? Maybe the VPS needs more "oomph" - excuse my highly technical language.
I will try on a non-VP machine.
But they are relatively fast and will probably make it. I'll share the result here.
Sure it is quite special to serve 2 crypto exchanges at the same time and to keep a total of 60 values up to date in the quote. And then all that on a VPS. I would even want to increase that if possible.
I would never let (half)automatic systems run on a "metal" at home.
QUOTE:
Lukas, could you tell us how the VPS is configured (CPU, RAM)?
Today the crypto markets are busier and my system is extremely slow again. I sometimes wait 20 seconds for an input.
On my fast physical PC (i9 8*2 cores, 4G) the load goes from about 3% to 20%.
So it is logical that my VPS (2*2 cores, 3G) then at its limit (100%).
What exactly occupies the performance?
On my fast physical PC (i9 8*2 cores, 4G) the load goes from about 3% to 20%.
So it is logical that my VPS (2*2 cores, 3G) then at its limit (100%).
What exactly occupies the performance?
We may have accidentally run into the cause just yesterday. If you're using Polling and not Streaming, the system is saving the result of the polling to [probable] large disk files. Doing this all at once puts a strain on the machine.
We're already testing "new lightweight polling" for the Strategy Monitor today and it seems to resolve the issue.
Meanwhile, you can switch to Streaming - as long as the provider supports the number of simultaneous streams required.
We're already testing "new lightweight polling" for the Strategy Monitor today and it seems to resolve the issue.
Meanwhile, you can switch to Streaming - as long as the provider supports the number of simultaneous streams required.
Your Response
Post
Edit Post
Login is required