A publication of the National Air Traffic Controllers Association
Issue link: http://natca.uberflip.com/i/775152
TFMS patch P6 was released on Jan 1 3th, Friday and P7 on the 14th. P6 was a regular patch that got delayed due to the system being busy and P7 was an emergency patch to fix a recent TFMS/FSM believed problem. One of the deliveries was for recent AFP failures, which has caused a lot of tu rmoil. A number of hypotheses have been st ated with no real pro ven solution from third party. Only suggested exploratory fixes have been the path so far. First it was believed an ECR change during a revision locked up everything the first time but that seemed to be false , which is the P7 fix. A nother was pointed to the way we were putting out the revision of the AFP ( A irspace Flow Program) taking to o long and not turning subs off. The testing at WJHC showed initially not turning off the sub did not cause the problem. It did not cause the error b ut more analysis is expected to come back by the 17th. Another fact showed up pointing to a dropped flight during a revision caused the system to lock up. This was a flight that was first in the AFP then during the revision, it left the AFP as a DO Droppe d flight. This testing is still happening at WHJC as of Jan 13th. It is believed these problems have always existed. It wasn ' t that noticeable since during ground delay programs and ground stops the automation process happens very quickly. However, in a revision of an AFP of 1000 flights or more , it can take up to two minutes , but on these day s, it took up to 7 - 10 minutes. This allowed the system to be pinged by many thin gs before completing its task. More inform ation will come out this week. This poi nts to the delicate problem surrounding TFMs and may continue until TFMS gets a real overhaul, not just patch work , which today is built with many fragments of language, some COTS software , and pieced systems to perform. The approach of adding bells and whi stles pressured by outside influences instead of building a sound integrated platform with well - supported follow - up enhancements will continue to make the system brittle. The new RAD tool, with ABRR/PDRR, will be the impact on the system Jan 23rd. It is unsure how the overall effect of the NAS can perform with it since a real system condition ca nnot be met in the current lab. They have problems at the lab just to do the current testing with bouncing dat a from one server to the next. On another note, t he Command Center is planning to run CTOP this summer, which can be built off a multitude of FCAs like AFPs with auto revisions causing massive refresh if that is the intent. My position is if they plan to do so, to do it a little at a time to explore the entire system ' s weak spots before pu rsing something overly complex. If anything comes from these unfortunate events and future strains like CTOP, it will be an action plan to fast - track a better TFMS platform with a better funded path for enhancing properl y.