Those missing segments


David J Taylor GM8ARV 🏴󠁧󠁢󠁳󠁣󠁴󠁿 🇪🇺
 

--- In SatSignal@yahoogroups.com, Lawrence <lawrence@a...> wrote:
I sat and watched the MSG-1 images build up today, whilst
monitoring the
CPU (etc) activity on the receive computer. Finally - several
images
lost a segment, but during that time the CPU was relaxing at a low
percentage. Whatever causes the segment loss on my computer, it is
*not* other programs starting up and hogging resources. I guess I
simply have to live with this.

Lawrence Harris
Lawrence, I don't believe it is my software that is causing the
problem, either. I will reply to you on the MSG-1 list.

Cheers,
David

I am still uncertain as to where the problem lies. I have a system
here that, under normal circumstances, doesn't loose segments. I
have seen a very occasional loss (say every other week) which I
could ascribe to the weekly defrag run on the disk with the Tellique
log files and staged MSG-1 data files. However, over the last few
days the system has been loosing a few segments a day, whereas a
second parallel system did not. This is similar to what other
people have reported - sudden unexplained losses of segments.

On checking the Tellique log file, I found that there were an awful
lot of "Missed file" messages around the period in question. I
understand this is normal, and refers to LRIT data which has been
interrupted by higher priority LRIT data, forcing the waiting files
to be transmitted later.

However, I did wonder whether all the activity of writing to the log
file with all these "missed file" messages was actually then causing
a data loss. Note that compared to Eumetsat's current standard I
run without the log file buffering enabled, and with verbose rather
than normal logging, so I may be creating problems for myself. I
have now turned on the log file buffered writing in the RECV.ini:

[locations] section: log_file =>> recv.log

So far today I ahve had one missing segment on the main PC, 1445,
HRIT ch 10, seg 4. The Transponder Status screen shows no
uncorrected blocks, and there is nohting like a bunch of missed file
messages in the Tellique log file. Why that segment is missing I
have no idea.

Suggestions:

- you may want to look at the log file for the period where you have
missing segments.

- you may want to run the Transponder Status screen looking at the
SNR and in particular for any uncorrected blocks. Warning, some
people have reported that this _causes_ missing segments.

- you may want the change the Tellique logging level and turn on
buffered log file writing.

- check if anything is blocking the satellite signal - aircraft,
birds, trees etc. I hear it is windy down south today - is your
dish absolutely aligned and steady?

Going back to my original comments about "missed files", I wonder
what it is that has to be given priority on the LRIT channel? I get
the impression that the LRIT channel is heavily loaded - could
Eumetsat purchase a little more bandwidth for it to reduce that
loading, and hence the congestion causing these file re-
transmissions. Just thinking aloud.....

Cheers,
David


Douglas Deans <douglas@...>
 

So far today I ahve had one missing segment on the main PC, 1445,
HRIT ch 10, seg 4. The Transponder Status screen shows no
uncorrected blocks, and there is nohting like a bunch of missed file
messages in the Tellique log file. Why that segment is missing I
have no idea.

Yes that is my only missing HRIT segment reported for today (since 07.30).
So probably not us.

Regards
Douglas.


a_van_belle
 

--- In MSG-1@yahoogroups.com, "Douglas Deans" <douglas@d...> wrote:

Yes that is my only missing HRIT segment reported for today (since
07.30).
So probably not us.
Same over here Douglas. Up to cycle 2030 this is my only missing
segment.
But I do have this "log_file=>>recv.log" entry in my recv.ini
Must have been somewhere in September that I read this tip on the
list and changed my original settings.

Greetings,
Arne van Belle