Support Board
Support Board
Search Board
Control Panel
View My Posts / Threads
Direct MessagesDate/Time: Fri, 03 May 2024 08:28:17 +0000
level 2 discrepancies..?
View Count: 1304
| [2016-05-19 15:54:58] |
| Cavalry3162 - Posts: 523 |
|
Hello support, recently i figured out that there are different volumes on the dom for iqfeed and your data feed.. even though a tick here and there can be lost, but those differences baffle me. are you sure your data is correct? can you somehow back this? i am aware that this might well be an iqfeed issue but it would be very nice of you when you could double check that on your end and / or with barchart directly. see attached screenshot, iqfeed in green and sc feed in blue. thanks, Andreas |
 2016-05-19_11-49-57.png / V - Attached On 2016-05-19 15:54:53 UTC - Size: 98.87 KB - 283 views |
| [2016-05-19 17:30:53] |
| |
|
Do you notice a problem at the moment? We would not expect that you would. If this problem occurs at the time there has been some packet loss, that would explain it. There can also be differences just simply due to subtle timing differences when those two Trade DOMs are updated. This is also a likely explanation. Some packet loss which is usually very small, does occur and we have been persistent with about Barchart about this. We plan to follow up with them again and be more forceful. We are getting very tired of it assuming this is the reason for the problem. Otherwise we do know the data is 100 percent correct as long as there is no packet loss. There was some packet loss for today during the 9 AM hour. For CBOT it was 0.002%. So you can see it is very small. So this means that there could have been a few momentary instances of this. But you should not notice any difference during any other hour today. At any other time today we are certain the data is 100 percent complete. Sierra Chart Support - Engineering Level Your definitive source for support. Other responses are from users. Try to keep your questions brief and to the point. Be aware of support policy: https://www.sierrachart.com/index.php?l=PostingInformation.php#GeneralInformation For the most reliable, advanced, and zero cost futures order routing, *change* to the Teton service: Sierra Chart Teton Futures Order Routing Date Time Of Last Edit: 2016-05-19 17:31:48
|
| [2016-05-19 18:52:20] |
| Cavalry3162 - Posts: 523 |
|
hello SC, unfortunately this still persists, see attached video. it isnt excatly a delay issue either, even in slow markets, the numbers are off. the issue is spread above multiple markets (i've tested CME (eMini S&P) and CBOT (Ten Year and 30 Year treasury future)) so, you think, this should be an IQFeed issue then? thank you for your help, Andreas |
Private File |
| [2016-05-19 19:05:38] |
| |
|
Not necessarily with IQ Feed because sadly we just saw some more incident of packet loss during the previous hour reported at 14:45 US Eastern time. This is unusual at this time. This is on channel 344 which has ZB. Total packets, packets lost, percentage lost: 633857, 36, 0.006 Although as you can see, this really is very small. So it would seem as though this would cause very sparse differences. Sierra Chart Support - Engineering Level Your definitive source for support. Other responses are from users. Try to keep your questions brief and to the point. Be aware of support policy: https://www.sierrachart.com/index.php?l=PostingInformation.php#GeneralInformation For the most reliable, advanced, and zero cost futures order routing, *change* to the Teton service: Sierra Chart Teton Futures Order Routing Date Time Of Last Edit: 2016-05-19 19:06:20
|
| [2016-05-31 07:56:12] |
| |
|
Regarding the minor packet loss we have seen (~.005%), this has been looked into more closely. The connection to our servers that receives the CME multicast feed now has been changed to a 10 gigabit connection which is the native speed of the switch. We also have come to realize that some of this issue could be happening on the server and not on the external network run by Barchart. We have made some changes and the above change has also been done by Barchart over the weekend. We will see how it goes today. Hopefully we will be down to 0% packet loss. Whether we have achieved that right now, or there is more to do, we will see. Also, our server updates the market depth levels from the CME snapshot feed every 15 minutes. If that is slightly lagged behind the incremental feed, that could account for some momentary differences as well. Not sure what we can do about that, because this periodic refreshing is important to ensure that the state of the data feed is always up-to-date in the event that there is data loss. Sierra Chart Support - Engineering Level Your definitive source for support. Other responses are from users. Try to keep your questions brief and to the point. Be aware of support policy: https://www.sierrachart.com/index.php?l=PostingInformation.php#GeneralInformation For the most reliable, advanced, and zero cost futures order routing, *change* to the Teton service: Sierra Chart Teton Futures Order Routing Date Time Of Last Edit: 2016-05-31 07:57:47
|
| [2016-05-31 16:34:08] |
| |
|
Letting you know, the changes that were made over the weekend did indeed take the packet loss to 0%.
Sierra Chart Support - Engineering Level Your definitive source for support. Other responses are from users. Try to keep your questions brief and to the point. Be aware of support policy: https://www.sierrachart.com/index.php?l=PostingInformation.php#GeneralInformation For the most reliable, advanced, and zero cost futures order routing, *change* to the Teton service: Sierra Chart Teton Futures Order Routing |
| [2016-05-31 19:11:56] |
| Cavalry3162 - Posts: 523 |
|
thanks for all this, i appreciate the effort to continuously improve your infrastructure and hone out any issues. there's still some minor differences at certain levels, just so you know, but its not really that i think this is a NoGo. Andreas |
Attachment Deleted. 2016-05-31_14-06-32.png / V - Attached On 2016-05-31 19:11:30 UTC - Size: 80.31 KB - 255 views |
| [2016-05-31 22:15:01] |
| |
|
We need to know at what time this occurs including the time zone because after we wrote what we did, we did see a small amount of packet loss occur in a later hour. As a percentage measurement, it was less than .001%
Sierra Chart Support - Engineering Level Your definitive source for support. Other responses are from users. Try to keep your questions brief and to the point. Be aware of support policy: https://www.sierrachart.com/index.php?l=PostingInformation.php#GeneralInformation For the most reliable, advanced, and zero cost futures order routing, *change* to the Teton service: Sierra Chart Teton Futures Order Routing Date Time Of Last Edit: 2016-05-31 22:15:27
|
| [2016-06-01 06:42:49] |
| Cavalry3162 - Posts: 523 |
|
its there all the time, no specific time or day..
|
| [2016-06-01 07:43:29] |
| Cavalry3162 - Posts: 523 |
|
made 2 new screenshots with time.. the offset seems "fix" as its the same amount between those 2 screenshots.
|
| 2016-06-01_03-40-10.png / V - Attached On 2016-06-01 07:43:22 UTC - Size: 100.14 KB - 227 views 2016-06-01_03-42-00.png / V - Attached On 2016-06-01 07:43:26 UTC - Size: 92.45 KB - 258 views |
| [2016-06-01 07:59:46] |
| Cavalry3162 - Posts: 523 |
|
and another one to make sure it persists even after your 15min reset
|
| 2016-06-01_03-58-40.png / V - Attached On 2016-06-01 07:59:42 UTC - Size: 104.95 KB - 273 views |
| [2016-06-01 08:11:29] |
| |
|
There is zero percent packet loss at this time, so that is not the cause of it. Knowing how the market depth data is processed and transmitted we are 100% confident the data is accurate. Even in the case when the most recent snapshot every 15 minutes from the CME is received and transmitted, even that is not a problem because that is up-to-date data direct from the CME and we would not expect it to be older than any more then maybe milliseconds. The other data feed you are comparing to does not look to have 100 percent up-to-date data at those levels. Also we are comparing to IQ Feed at the moment for the ES June 2016 futures. We do not see any differences. They are an identical match after monitoring for a couple of minutes. When the values are updating frequently, they can be different for short moments in time, but once a quantity settles for a second or two, we see a match. Sierra Chart Support - Engineering Level Your definitive source for support. Other responses are from users. Try to keep your questions brief and to the point. Be aware of support policy: https://www.sierrachart.com/index.php?l=PostingInformation.php#GeneralInformation For the most reliable, advanced, and zero cost futures order routing, *change* to the Teton service: Sierra Chart Teton Futures Order Routing |
| [2016-06-01 08:13:26] |
| |
|
In response to post #11 we are confident the data you are seeing from the Sierra Chart Exchange Data Feed is accurate. With everything we know we have no reason to think otherwise. Sierra Chart Support - Engineering Level Your definitive source for support. Other responses are from users. Try to keep your questions brief and to the point. Be aware of support policy: https://www.sierrachart.com/index.php?l=PostingInformation.php#GeneralInformation For the most reliable, advanced, and zero cost futures order routing, *change* to the Teton service: Sierra Chart Teton Futures Order Routing |
| [2016-06-01 08:23:21] |
| Cavalry3162 - Posts: 523 |
|
Hello Support, the "other feed" is iqfeed.. however, thank you for checking this! I verified that the ES is 100% identical, its the treasuries which dont be the same. (made another screenshot). so, i guess its an iqfeed issue then even though its strange that its always higher numbers on iqfeed. i understand that there are some pulls but if its always a higher number its odd.. Andreas |
| 2016-06-01_04-17-01.png / V - Attached On 2016-06-01 08:23:16 UTC - Size: 256.89 KB - 226 views |
| [2016-06-02 06:16:58] |
| |
|
We have looked into this, and we are 100 percent certain the data you are getting from the Sierra Chart Exchange Data Feed is correct and is exactly what the CME is providing. The way that the market depth data is processed for the ES is identical to all other symbols. So if it is correct for that symbol, it is correct for other symbols. Perhaps it is possible, although we have never had any clear indication of this, that the periodic refresh of the depth data from the snapshot data every 15 minutes could under some conditions possibly cause a momentary issue with the market depth data. After looking over the screenshots you provided, we do not think this applies here. It is also possible with the way the CME transmits market depth updates on the incremental feed, there may not be any possible issue to begin with. That is probably what we would expect. Sierra Chart Support - Engineering Level Your definitive source for support. Other responses are from users. Try to keep your questions brief and to the point. Be aware of support policy: https://www.sierrachart.com/index.php?l=PostingInformation.php#GeneralInformation For the most reliable, advanced, and zero cost futures order routing, *change* to the Teton service: Sierra Chart Teton Futures Order Routing Date Time Of Last Edit: 2016-06-02 06:18:01
|
| [2016-06-02 06:55:40] |
| Cavalry3162 - Posts: 523 |
|
alright, thanks for taking the time to look into this again. like i said, i dont think its a real issue anymore, the latest posts were more of a fyi. a couple of contracts here and there dont really matter. it is interesting, though, that those kind of issues exist in the first place. i would not have thought that these could happen (whoever to blame for, thats not my point). thanks again, Andreas |
To post a message in this thread, you need to log in with your Sierra Chart account: