1

SADISOPSG/14-IP/10
/
International Civil Aviation Organization
WORKING PAPER / SADISOPSG/14-IP/10
19/5/09

SATELLITE DISTRIBUTION SYSTEM OPERATIONS GROUP (SADISOPSG)

FOURTEENTH MEETING

Bangkok, Thailand, 15 to 17 July 2009

Agenda Item / 6: / Development of the SADIS
6.3: / Report of the SADISOPSG Technical Developments Team

STATISTIC INFORMATION ABOUT DATA LOSSES

(Presented by the SADISOPSG Technical Development Team)

SUMMARY
This paper presents some statistics about data losses on SADIS and shows the difficulties at present to make comparable statistics.

1.INTRODUCTION

1.1During the last few SADISOPSG meetings there were some discussions on data losses on the user side and also a paper was presented concerning these problems of data loss. Since then some action has been undertaken by different parties from the provider side and resulted in some improvement, but data losses still happen.

1.2A very good example of the users view is also by comparison of the questionnaire of efficacy, where many user stated no loss of data and in WP/19 from the provider State is mentioned:

a)MeteoSwiss reports in December 2008 and January 2009 suggested that data loss (drop out) was occurring at around the same time each day (1505 UTC) on the OPMET channel. A loss of synchronisation was also apparent on the Met Office SADIS 2G Comparator system, which corroborated the MeteoSwiss reports.

It is not clear why such a long period of drop outs are not reported or observed from other users.

1.3This paper presents background information about the time period March/April 2009 of missing data or unknown receiving status at six different geographical sites with the same result and are checked against the results at MCH in Zurich. The comparison sites are located in South Africa (SWAS), Belgium (Belgocontrol), Sweden (LVF), Lufthansa Systems (Frankfurt and Gdansk) and Pretoria (Netsys). The equipment used at these sites where supplied by LTeq, Paradigm and NetSys and the installation is done by different supplier.

1.4The statistics are based on jumps in sequence numbers or sequence errors for all of the three channels OPMET, T4 and GRIB. The reason for these errors are still not clear, number jump backwards, gaps in sequence numbers or sometime no message received for 10 minutes. The transmission problem does not necessarily occur on all channels at the same time.

1.5To make a clear statistic a continuous sequence number flow is needed. Until now this is the only way to check the completeness of the OPMET data.

1.6In case of OPMET data losses for any reason there is no way to identify the data which were lost. The only place where the completeness of the data transmission can be checked is at the provider side.

1.7It should be useful if the users should be able to compare the data on e.g. the FTP server if that data is a mirror from the uplinked data including the same sequence number.

1.8In addition Zurich installed some automatic checks in July 2008 for the case of no OPMET data reception for 10 minutes and automatically sends service messages of the following content to NATS:

BMBB41 LSSW 071550

GG EGGGYF..

071550 LSSWYM..

Information for NATS:

No SADIS_OPMET Input (SADIS2G) in the last 10 minutes.

Reset of all SADIS processes on our site done!

For questions, please contact:

...... @meteoswiss.ch or phone +41(44)256 .... =

1.9Due to misinterpretation of such messagesat NATS no action has been taken, but by writing this IP the procedures have been changed in order to take appropriate action on receiving these messages. At the time writing this paper, this procedure seems to be successful.

2.Summary of Statistics in the period March and April 2009

OPMET Data / T4 Data (BUFR) / GRIB Data
Period of time / Count days with data loss / Count bulletins lost / Count days with data loss / Count bulletins lost / Count days with data loss / Count bulletins lost
March 2009 / 5 / * / 1 / * / 3 / *
April 2009 / 4 / ** / 0 / 0 / 1 / 15
Legend
*a clear statement not possible; sequence number wrapped out / **a clear statement not possible, >100; sometimes sequence number jumps backwards

3.Detailed information

OPMET:

Thu Mar 12 09:31:46 942 SAHK31 VHHH 120930 nul

Missing 943 to 954

Thu Mar 12 09:33:08 955 FTCN36 CWAO 120900 AAA

------

Thu Mar 12 20:08:09592 SARE20 FMEE 122000 nul

Missing 593 to 591

Thu Mar 12 20:08:30 592 FCEW31 LEMM 121700 AAA

------

Sun Mar 15 21:13:52 794 SAUS59 EGGY 152100 RRA

Missing 795 to 796

Sun Mar 15 21:13:52 797 SAUS63 EGGY 152100 RRA

------

Fri Mar 20 15:08:11 676 SAUS60 EGGY 201500 RRA

Missing 677 to 679

Fri Mar 20 15:08:11 680 SPCN42 CWAO 201504 nul

------

Sat Mar 21 20:54:39 505 SAMA20 FIMP 212100 nul

Missing 506 to 507

Sat Mar 21 20:58:28 508 WSPA10 PHFO 212052 nul

------

Fri Apr 17 14:55:24 191 SANO35 ENMI 171450 nul

Missing 192

Fri Apr 17 14:55:24 193 SADL35 EDZO 171450 nul

------

Fri Apr 17 14:55:24 193 SADL35 EDZO 171450 nul

Missing 194

Fri Apr 17 14:55:24 195 SANO38 ENMI 171450 nul

------

Fri Apr 17 14:55:24 195 SANO38 ENMI 171450 nul

Missing 196

Fri Apr 17 14:55:24 197 SAPM31 MPTO 171500 nul

------

Fri Apr 17 14:56:36 216 SAAO33 DRRN 171400 RRA

Missing 217 to 218

Fri Apr 17 14:56:39219 SAUK33 EGGY 171450 RRA

------

And many more between 14:56:46 and 15:04:35 on April 17

------

Sat Apr 18 15:06:26 727 SAJP42 KWBC 181500 nul

Missing 728 to 729

Sat Apr 18 15:06:26 730 SASA31 KWBC 181500 nul

------

Sat Apr 18 15:06:29 744 SAUS63 EGGY 181500 RRA

Missing 745 to 747

Sat Apr 18 15:06:29 748 SAUS67 EGGY 181500 RRA

------

Sun Apr 19 15:05:32 786 SACN31 CWAO 191500 nul

Missing 787 to 788

Sun Apr 19 15:05:32 789 SAGU31 MGGT 191500 nul

------

Tue Apr 21 15:15:22 765 FTCN36 CWAO 211500 AAA

Missing 766

Tue Apr 21 15:15:22 767 FCEW34 LEMM 211400 AAA

------

Tue Apr 21 15:28:56 899 SPCN42 CWAO 211527 nul

Missing 900

Tue Apr 21 15:30:38 901 WSRH31 LDZM 211530 nul

------

And many more between 15:56:12 and 17:46:07 on April21

SADIST4:

Fri Mar 20 20:31:25 757 PGJE05 KKCI 201200 nul

Missing 758 to 999

Fri Mar 20 23:04:02 000 PFXD05 ADRM 202300 nul

SADISGR:

Thu Mar 12 09:31:55 162 HUOE10 EGRR 120600 nul

Missing 163 to 287

Thu Mar 12 09:33:04 288 HVME96 EGRR 120600 nul

------

Fri Mar 20 15:24:05 994 HHOG20 EGRR 201200 nul

Missing 995 to 996

Fri Mar 20 15:24:08 997 HHJG20 EGRR 201200 nul

------

Fri Mar 20 15:28:09 407 HHLF97 EGRR 201200 nul

Missing 408 to 999

Fri Mar 20 23:42:18 000 HTMB85 EGRR 201800 nul

Tue Apr 21 15:29:02 897 HHNF85 EGRR 211200 nul

Missing 898 to 913

Tue Apr 21 15:29:35 914 HHOF60 EGRR 211200 nul

4.ACTION BY THE SADISOPSG

4.1The SADISOPSG is invited to note the content of this paper.

— END —