David J Taylor GM8ARV 🏴 🇪🇺
David,`EUMETSAT Data Channel 2', address 18.104.22.168:2511 (subscribed)
VRB:2003-10-30 11:48:15.109:Connecting to data channel `EUMETSATData Channel 2', address 22.214.171.124:2511 (invited)
MSG:2003-10-30 11:48:15.437:Connect to data channel `EUMETSAT DataChannel 2', address 126.96.36.199:2511 completed (invited)
VRB:2003-10-30 11:48:16.484:Received filelist 3fa0faa10009f919 onchannel `EUMETSAT Data Channel 2'
MSG:2003-10-30 11:48:16.531:File transmission 3fa0f95e0009f850ended/interrupted: 1 file missing/incomplete
VRB:2003-10-30 11:48:16.687:Received new key 6526
OK, well when checking my log for the same period, I also missed
part of a file:
VRB:2003-10-30 11:48:50.826:Connecting to data channel `EUMETSAT
Data Channel 2', address 188.8.131.52:2511 (invited)
MSG:2003-10-30 11:48:51.147:Connect to data channel `EUMETSAT Data
Channel 2', address 184.108.40.206:2511 completed (invited)
VRB:2003-10-30 11:48:52.218:Received filelist 3fa0faa10009f919 on
channel `EUMETSAT Data Channel 2'
MSG:2003-10-30 11:48:52.248:File transmission 3fa0f95e0009f850
ended/interrupted: 1 file missing/incomplete
MSG:2003-10-30 11:48:52.278:Missed parts of file `L-000-MSG1__-
3fa0f95e0009f84f from channel `EUMETSAT Data Channel 3'
Differences? I have Tellique logging set to verbose, and I am
running the V2.3.1 (beta) client. My own program did not detect
that part of the segment set was missing, so what was supposedly an
incomplete file had been delivered by the Tellique software to the
\received folder. Interesting! Looking at that file with an
analyser also suggests that it is complete (the equivalent file
length matches the expected data length - 600,000 bytes), however,
there is a sequence error in that CLM______-000005 arrived after
CLM______-000006. Is this significant? (I hope Eumetsat are
reading this!). This appeared to come between HRV segment 5 and
VIS006 segment 2. Of course, the HRV and VIS006 were transmitted on
data channel 2 (HRIT), and the MPEF file on data channel 3 (LRIT).
The other thing I checked was the file size. With the wavelet-
compressed JPEG data, the segment files are now much larger.
However, it seems to be segment 10 which is the largest, rather than
segments 16 and 17 which are causing David S's problem.
Neither of you said precisely when this problem started, by the way.
Luca's comment is a valid one - we may expect some random errors due
to solar activity right now.
Still looking for ideas...