Is IB making our orders & quotes compete with back-fill/charting bandwidth?

Quote from IBsoft:

We turned the charting back on. You should not see either the market-data or order-ack times having deteriorated as a result.

i can confirm this, acks are still nice and tight today. thanks guys
 
electron, I had the same problem. Try this: Go into your TWS directory, there's a folder starts with D, in there delete "settings.xml". Then start up TWS, and it will give you the default window, which you can go reset by going into file > manage wprkspaces > pre-upgrade
 
Quote from gms:

electron, I had the same problem. Try this: Go into your TWS directory, there's a folder starts with D, in there delete "settings.xml". Then start up TWS, and it will give you the default window, which you can go reset by going into file > manage wprkspaces > pre-upgrade

We are working on making the settings savings more reliable.
 
Quote from Quartz:

Thanks IBSoft - appreciate reading the last reply you gave above.

Yes, I concur. IB catches some flack here at times, but I'm not aware of any broker that is more responsive, more on top of issues and more accessible to users.
 
Quote from IBsoft:

The historical charting (a.k.a. backfills) and scanner functionality is implemented as an entirely independent system, consisting of separate processes running on different pieces of hardware and having its own vast data stores.

It shares one element w/ the order-routing and market-data subsystems: the firewalls. Without giving you specific data amounts (so as not to give our competitors a hint about our data-rates and compression ratios) I will reveal to you three relevant things:
- the total charting and scanner data we send out is less than 1.5% of the real-time market data we send out
- the firewalls have plenty of capacity on them, i.e. they can sustain several times higher loads that we experience both on average and in peak times
- the excess capacity of the firewalls is continuously monitored; when it falls below a (generous) threshold, we simply add more of them - it is a well scalable setup

The reason why the firewalls are shared between the systems is simple. Great many customers of ours are behind their own or corporate firewalls. We do not want them to have to further open up their firewalls when we add a service, or a data-farm or something.

We turned the charting back on. You should not see either the market-data or order-ack times having deteriorated as a result.
Thank you Ib for this clarification. I believe that this puts to rest some legitimate concerns of many of us. It would perhaps have been preferable to have known about this a bit earlier.

One point that seems clear to me is the following:
Backfill and to some extent charting enables the interested users to catch up with the current state of the market.

Trading customers have an even much greater urge to 'catch up' and re-establish connection in case of an accidental interruption, whatever the cause, Ib's ISP, a misconfigured firewall, customer's ISP, you name it. This concern MUST have Ib's priority given the financial risk the customer is carrying by virtue of his possible open positions. As an automated API user, I can tell you that this is a most difficult problem to solve given the present lack of true reconnect support in TWS. Please make it your first priority to come up with a foolproof solution. The 'freeze' problem, selective or not, is very real. Simply read your customer's posts of the last few months. You may shrug this off as insignificant, but this ain't so from a customer's viewpoint.

Thank you Ib,
nononsense
 
Quote from nononsense:

Thank you Ib for this clarification. I believe that this puts to rest some legitimate concerns of many of us. It would perhaps have been preferable to have known about this quite in advance.

One point that seems clear to me is the following:
Backfill and to some extent charting enables the interested users to catch up with the current state of the market.

Trading customers have an even much greater urge to 'catch up' and re-establish connection in case of an accidental interruption, whatever the cause, Ib's ISP, a misconfigured firewall, customer's ISP, you name it. This concern MUST have Ib's priority given the financial risk the customer is carrying by virtue of his possible open positions. As an API user, I can tell you that this is a most difficult problem to solve given the present lack of true reconnect support in TWS. Please make it your first priority to come up with a foolproof solution. The 'freeze' problem, selective or not is very real. Simply read your customer's posts of the last few months.

Thank you Ib,
nononsense

This weeks two mishaps (the misconfigured firewall and our ISP failure) have shown that our reconnect policy, while ok when a few hundreds or thousands TWSs need to reconnect, can be improved for cases when all our users need to reconnect en-mass. This MUST and DOES have the highest priority. The enhancements are already being tested as we speak, the rollout will however take a bit of time.

At the same time
- we are making an API upgrade that will make the connection status to the various services transparent
- the loss of settings, which coincided w/ this week connectivity loss is now understood and being rectified

As to the freeze problem, I have been following the discussions in the forum. We have worked closely with some of posters and addressed all the (to us) known issues. I need to add that some of the posters are in countries with weak ISP services. There is a limit to how stable the TWS can be in such circumstances. I have attempted to explain the basics of the socket connections in one or my earlier posts so as to point out that it is not that we are ignoring those users, but that there is a limit as to how much we can do for them.
 
nononsense,
We never shrugged off the "freeze" problem. In fact I even PM'd you on 6/27 asking for specific information if you were having the problem. I never heard back from you.

Quote from nononsense:

Thank you Ib for this clarification. I believe that this puts to rest some legitimate concerns of many of us. It would perhaps have been preferable to have known about this a bit earlier.

One point that seems clear to me is the following:
Backfill and to some extent charting enables the interested users to catch up with the current state of the market.

Trading customers have an even much greater urge to 'catch up' and re-establish connection in case of an accidental interruption, whatever the cause, Ib's ISP, a misconfigured firewall, customer's ISP, you name it. This concern MUST have Ib's priority given the financial risk the customer is carrying by virtue of his possible open positions. As an automated API user, I can tell you that this is a most difficult problem to solve given the present lack of true reconnect support in TWS. Please make it your first priority to come up with a foolproof solution. The 'freeze' problem, selective or not, is very real. Simply read your customer's posts of the last few months. You may shrug this off as insignificant, but this ain't so from a customer's viewpoint.

Thank you Ib,
nononsense
 
Quote from gms:

electron, I had the same problem. Try this: Go into your TWS directory, there's a folder starts with D, in there delete "settings.xml". Then start up TWS, and it will give you the default window, which you can go reset by going into file > manage wprkspaces > pre-upgrade

I was told the same thing by the IB support and the problem was fixed. Thanks, anyway...
 
Quote from samaritan:

nononsense,
We never shrugged off the "freeze" problem. In fact I even PM'd you on 6/27 asking for specific information if you were having the problem. I never heard back from you.
True. Like others have already indicated, I don't think that ET identity should be linked to Ib identity. Besides, for many months you gave the distinct impression not to take things very seriously. In fact, a few minutes ago you talked for the first time about a 'freeze' problem, which you kind of want to blame 'weak' ISP's for. How do you then explain the 'selective' freeze phenomenon. Can't be a 'weak' ISP. Reprimanding a customer not wanting to foresake his forum privacy ain't going to help much.

This may help though:
http://www.interactivebrokers.com/discus/messages/3/29970.html?1123247522
http://www.interactivebrokers.com/discus/messages/2/29972.html?1123185255

Thank you samaritan for your most thoughtful concern.
 
Back
Top