Today's VPS update:
Running 7 strategies, all automated. Everything ran fine. I'm living the dream.

For your amusement....

CPU and Memory Usage


I closed and reopened RDP several times, but I did not experience the shutdown of TS that I had Monday or Friday. I will continue to try to replicate that issue.

~vmodus

Prophet85 

I suggest downloading PingPlotter which is free to use.

Big fan of pingplotter. So much so I actually pay the $10/year maintenance for it. Great product.

https://www.pingplotter.com/

*Not an advertisement. No relationship to them. Have used the software for just under 7 years.

Interesting test: Using pingplotter

trade.tt gave me 21 ms RT

m.cqg.com gave me 41 ms RT

google.com gave me 23 ms RT

What I really want to see whats the hedge latency for my TT setup for an autospreader and the latency between where i am located in TT to the exchange on outright trades

Quick update: simple ping test to trade.tt from across the Globe was 200 ms several times (some were comparable to the US) - so justifies VPC, automation etc even on hosted platforms


Sent using the NexusFi mobile app

jokertrader 

Interesting test: Using pingplotter trade.tt gave me 21 ms RT m.cqg.com gave me 41 ms RT google.com gave me 23 ms RT What I really want to see whats the hedge latency for my TT setup for an autospreader and the latency between where i am located in TT to the exchange on outright trades

If you are interested, using old fashioned command line ping:

• trade.tt >> 14ms from my office; 8ms from my VM
• m.cqg.com >> 35ms from my office; 12ms from my VM

~vmodus

Our VM is still running strong, now with 9 automated strategies on TradeStation, averaging 115 round trip entries per day. Averaging 11.3% CPU utilization and memory utilization has been steady at about 50% (4 GB out of 8 GB).



I have also been doing some additional analysis while the strategies have run (new charts, new strategies, performance reports, exports), which you will see as the occasional spikes (also installed Office). It is running just fine.

As it is, this is my last update, unless anyone has any questions. The VM passes our tests and we will use live next week, once we add some additional security. I'm very happy with the results.



~vmodus

jokertrader 

Interesting test: Using pingplotter trade.tt gave me 21 ms RT

trade.tt is their website and not the IP address of their trading infrastructure - or at least the cloud front end. Unfortunately they do not have ICMP (ping) enabled on their network so you can't ping them. Pingplotter does return the route, but doesn't have the time for the last hop.

jokertrader 

What I really want to see whats the hedge latency for my TT setup for an autospreader and the latency between where i am located in TT to the exchange on outright trades

I believe this is how NextGen TT works.
1/ You have your browser, mobile or desktop and it communicates with TT over the internet.
2/ TTs outward facing infrastructure is cloud based (AWS I think). This is where things like the order book etc are all maintained. It is the middle-ware between you and the trading engines.
3/ The Trading engines are co-located at the exchange.

So if you send an order to buy X, it crosses the internet, hits their cloud structure, then goes to the trading engine. But if you launch an ADL Algo or Autospreader, you send that instruction to their cloud structure, which in turn sends the instruction to the trading engines, BUT the algo actually runs on the trading engine, and does not interact with the cloud structure other than to report relevant information. So the latency on an autospreader is tiny. If you look at your Audit Trail you can see what the claimed latency is. Having been on NextGen TT fully now for several months I am confident that it is substantially faster than my co-located XTrader servers were.

jokertrader 

Quick update: simple ping test to trade.tt from across the Globe was 200 ms several times (some were comparable to the US) - so justifies VPC, automation etc even on hosted platforms

If your outside the US and trading on TT I believe they have alternative on ramps to their network that should be quicker than going across the public network.

vmodus 

As it is, this is my last update, unless anyone has any questions. The VM passes our tests and we will use live next week, once we add some additional security. I'm very happy with the results.

Slightly off topic but are you having an trade issues? When I was running Tradestation on an 'AWS Desktop' there were mornings that as soon as I logged on Tradestation would execute a trade and when I checked it should have been executed at some point during the night. It was as if the server or Tradestation had gone to sleep and the process of me logging onto the box woke it up. Didn't happen all the time, but it happened enough to be considered an onging problem.

I'd be interested in hearing an update from you again after you're done with the security on which security route you chose.

vmodus 

Our VM is still running strong, now with 9 automated strategies on TradeStation, averaging 115 round trip entries per day. Averaging 11.3% CPU utilization and memory utilization has been steady at about 50% (4 GB out of 8 GB). I have also been doing some additional analysis while the strategies have run (new charts, new strategies, performance reports, exports), which you will see as the occasional spikes (also installed Office). It is running just fine. As it is, this is my last update, unless anyone has any questions. The VM passes our tests and we will use live next week, once we add some additional security. I'm very happy with the results. ~vmodus

SMCJB 

Slightly off topic but are you having an trade issues? When I was running Tradestation on an 'AWS Desktop' there were mornings that as soon as I logged on Tradestation would execute a trade and when I checked it should have been executed at some point during the night. It was as if the server or Tradestation had gone to sleep and the process of me logging onto the box woke it up. Didn't happen all the time, but it happened enough to be considered an onging problem.

There were no issues that I could see. There were no gaps in transactions and no stuck or pending orders. However, my process may be a little different. Here is what I do:

Sunday night

• Start VM
• Launch TradeStation with my trading desktop
• Check parameters for all strategies and instruments
• Check strategy automation settings
• Let it run

Each evening after close:

• Check for any problems
• Check my reports and look at any interesting metrics

Friday afternoon:

• Check to see if my strategies exit their positions at the assigned time
• Verify that no positions are open by the close
• Generate reports, etc.
• Shut down the VM for the weekend

So the VM is never down during the trading week. I can monitor the VM from my Azure dashboard. I usually keep the RDP session running in the background on my PC, but sometimes I just disconnect and let it run. I actually power down my regular PC almost every night now.

~vmodus

planetkill 

I'd be interested in hearing an update from you again after you're done with the security on which security route you chose.

Will do.

~vmodus