|
PureBytes Links
Trading Reference Links
|
I have noticed this happening. One item to beware of with the ROS fix is
the previously discussed "spontaneous ROS load" conflict with TS4. Not
everyone experiences it, but some of us have yet to make the new ROS work
successfully. Basically, in the middle of the trading day the server will
stop collecting data and start loading the ROS into the receiver. I tried
a number of fixes and finally gave up. (Who cares about dow quotes over
10,400 anyway? ) Again, not everyone experiences this, but it is quite
frustrating if you do.
CC
-----Original Message-----
From: Larry Wright <lwright@xxxxxxxxxx>
To: Omega List <omega-list@xxxxxxxxxx>
Date: Thursday, May 27, 1999 11:03 AM
Subject: Possible Signal data problem fix
>There is a potential problem with the Signal data feed. One symptom is a
>server message of "Quote for WP M9/UNDEF being stored in WP M9/CME" or
>something similar. Data is *not* actually stored, but is usually discarded
>(but not always). I found that data does show up on non-trading days,
>which alerted me to watch for other errors.
>
>Another problem is tick charts that 'reset' during the day, and will not
>plot earlier data even if the data is refreshed.
>
>I did some digging, and found the error ocurred when the exchange field in
>the Signal data was blank. Here are some selected data records within a
>one minute period (chopped a bit so you can see the end of the records in
>email):
>
>/ACa,9,ES M9,134100,,,137600,137450,137350,,,F@@@@,H,M,156,0,,,,M,846
> (note this record has an "M" in the next to last field; M = CME exchange)
>/ACa,9,ES U9,135475,135275,135550,138850,138650,,,GB@@@,H,M,38,0,,,,,846
> (note this record has NOTHING in the next to last field[last is time
field])
>/ACa,9,ES M9,134100,134075,137450,137350,,,G@@@@,H,M,156,0,,,,M,846
>/ACa,9,ES U9,135475,135250,138850,138650,,,GB@@@,H,M,38,0,,,,,846
>/ACa,9,ES M9,134100,134100,137450,137350,,,G@@@@,H,M,156,0,,,,M,846
>/ACa,9,ES U9,135475,135300,138850,138650,,,GB@@@,H,M,38,0,,,,,846
>/ACa,9,ES M9,134100,134100,137450,137350,,,G@@@@,H,M,156,0,,,,M,846
>/ACa,9,ES U9,135475,135325,138850,138650,,,GB@@@,H,M,38,0,,,,,846
>
>I verified that this was not just a TS server quirk by looking directly at
>the ASCII from the Signal box - it showed the same problem. The signal box
>was the source.
>
>The server does not seem to handle the missing fields correctly.
>
>I first called Omega, and after a series of emails back and forth, with
>examples and even a possible explanation, they had no solution. There was
>also no apparent interest in working on it, and the inquiry "died" with no
>useful outcome.
>
>I then called DBC and after the usual looonnnngggg wait, they didn't know
>what was happening either. I seemed to know more about the problem than
>they did, and insisted on talking with someone more versed in the details.
>The most experienced tech I could talk with could not discuss the problem
>- he did not know how data is actually transmitted. I even faxed the above
>data to them. He did offer to look into it, though.
>
>In a few days, I got an email asking about the ROS version - seems they
>have upgraded since the TS4 release. I downloaded the new ROS, version
>75.05 and it seems to help. It has cured the "Quote for WP M9/UNDEF being
>stored in WP M9/CME" problem, but has NOT fixed the tick chart reset
>problem. Hey - one out of two is not bad :-).
>
>Check your ROS version in the server. If it is an older one, download the
>75.05 ROS update from the DBC web site and install it. It might help with
>some problems, at least.
>
>Larry Wright
>
|