TelliCastStatsCollector.cmd 3.1 Beta


Ernst Lobsiger
 

Dear All,

here is Arne's and David's TelliCastStatsCollector.cmd 3.1 Beta for Windows polished. Needs no external program or library anymore. All done with some PowerShell magic. Must be scheduled minutes 29 and 59 every hour.

Please Beta test after rename *.txt-->*.cmd and report on MSG-1.

Cheers,
Ernst


Alan Curnow
 

Ernst,

I scheduled it to run as specified and at 23:59 it produced a file giving just 1 line of the date and time.
MyIp has been set to be the same as the Interface Address listed in cast-client_bas.ini
Is Terra the name of the receiving computer and is only used for the filename? 

Reading David's web page it says Port must be 8100 for Bas so I changed it to that and it produced another line with date, time and several numbers. It looks like at 00:29 it will add some headers for the following day.

Here's the output at 00:29 though it's actually 23:29 UTC so no headers yet. :)
The first line was port 8000, while line 2 was run manually with port 8100 which looked better.

Cheers,
Alan


Ernst Lobsiger
 

Alan,

so far so good. Node is only computer name in the header. The header is only produced when a new file is started. It will explain the numbers. You should not reset the statistics page of your TC client manually because this is where the numbers come from. I could only test here over the network to a GNU/Linux client but it seems to work with the Windows client as expected. See tomorrow what you have got so far ...

Cheers,
Ernst


Alan Curnow
 

Ernst,

I tried running it from a batch file using start to run minimized and low priority

start "STATS" /min /low C:\EUMETSAT\Utils\TelliCastStatsCollector.cmd

TelliCastStatsCollector runs fine but the batch file appears to open two minimized cmd windows on top of each other and when TelliCastStatsCollector finishes one of them closes. The other has to be closed manually.
Adding exit as a second line to the batch file did nothing nor did adding CMD /C to the start of the line. Can you see what I'm doing wrong?

Cheers,
Alan


Ernst Lobsiger
 

On Wed, Oct 20, 2021 at 05:25 PM, Alan Curnow wrote:
start "STATS" /min /low C:\EUMETSAT\Utils\TelliCastStatsCollector.cmd
Alan,

I havn't scheduled that under Windows 10 here yet. But the point might be that TClogSummary checks how it was called (because you can also use it interactively) and does an extra exit when started while TelliCastStatsCollector does not (because it is meant to be scheduled only). If you START a batch file an extra CMD with /K is started. So you must try to close that at the end: Add     ^& exit     to the end of your command line in the scheduler. That should do the trick.

start "STATS" /min /low C:\EUMETSAT\Utils\TelliCastStatsCollector.cmd ^& exit

Hope this works,
Ernst


David J Taylor GM8ARV 🏴󠁧󠁢󠁳󠁣󠁴󠁿 🇪🇺
 

On 20/10/2021 20:33, Ernst Lobsiger via groups.io wrote:
here is Arne's and David's TelliCastStatsCollector.cmd 3.1 Beta for Windows
polished. Needs no external program or library anymore. All done with some
PowerShell magic. Must be scheduled minutes 29 and 59 every hour.
I schedule my scripts for 00:29:01 on <some date when first created> and repeat
every 30 minutes indefinitely. No need for two entries.

When resetting the statistics I actually do that three times, although two may
be enough. I didn't trust doing it just once!

Cheers,
David
--
SatSignal Software - Quality software for you
Web: https://www.satsignal.eu
Email: david-taylor@blueyonder.co.uk
Twitter: @gm8arv


Alan Curnow
 

On Thu, Oct 21, 2021 at 07:27 AM, Ernst Lobsiger wrote:
start "STATS" /min /low C:\EUMETSAT\Utils\TelliCastStatsCollector.cmd ^& exit
Yes, that works Ernst. Thanks.

Alan


Alan Curnow
 

Ernst,

Would it be useful to put the service monitored in the filename, determined perhaps by Port number, so that stats can be collected on other services too without having to create separate subdirectories for them.

Cheers,
Alan


Ernst Lobsiger
 

On Thu, Oct 21, 2021 at 06:40 AM, Alan Curnow wrote:
On Thu, Oct 21, 2021 at 07:27 AM, Ernst Lobsiger wrote:
start "STATS" /min /low C:\EUMETSAT\Utils\TelliCastStatsCollector.cmd ^& exit
Yes, that works Ernst. Thanks.

Alan
Alan, Graham, David and All,

here comes TelliCastStatsCollector version 3.2 with all improvements:

- It does now auto exit when started,  ^& exit is not needed anymore
- it does now open a non existing file with writing the header first
- it makes a connection check to the client (writes ERROR entries)
- it makes 2 midnight resets as proposed by David in a recent post
- it commes a TelliCastStatsCollector0/1/2 for BASIC/HVS-1/HVS-2

Should be out of Beta now. We'll see ...

Cheers,
Ernst


Alan Curnow
 

Thanks Ernst,

Now ran the following batch file triggered to start at 16:59:01 (Local) every day to run every 30mins for 1 day.

start "BASIC" /min /low C:\EUMETCast\TelliCastStatsCollector0.cmd
start "HVS-1" /min /low C:\EUMETCast\TelliCastStatsCollector1.cmd
start "HVS-2" /min /low C:\EUMETCast\TelliCastStatsCollector2.cmd

Here's the results
Forgot to change the equipment description lines and number delimiter, will now do that.
The three separate minimized windows ran simultaneously and all exited correctly.

Cheers,
Alan
 


Ernst Lobsiger
 

On Thu, Oct 21, 2021 at 09:03 AM, Alan Curnow wrote:
run every 30mins for 1 day.
Alan,

you can even run it idefinitely as David proposed. One thing I see you are 1 hour off with respect to UTC. There is no problem as long as local time is an integer number of hours off.
Scheduling UTC with the Windows task scheduler is a nightmare. Nobody realy understands how that works. I always run my PCs under UTC. Makes things easier for an old brain ...

Cheers,
Ernst


Alan Curnow
 

Ernst,

Local time here is always 1 hour ahead or equal to UTC so it hopefully that shouldn't be a problem. The receiving PC doesn't do anything else but Tellicast so I could change it to UTC I suppose. 

Does setting a scheduled task to run indefinitely from a specific time continue after a computer restart? Presumably it does if that's what David does.

I had my 10min ramdisk cleanup to recur daily from a specific time and repeat every 10 mins, but on rebooting it stopped working, presumably until that time occurred again to trigger it, and the ramdisk filled up. I therefore set the ramdisk cleanup to be triggered by a restart and to occur indefinitely which works, but of course the specific 10 min start can't be set which in this instance doesn't matter. 

Cheers,
Alan


Graham Woolf
 

Hi Ernst

Got mine running - no problems so far

Regards

Graham


Ernst Lobsiger
 

On Thu, Oct 21, 2021 at 10:49 AM, Graham Woolf wrote:
Got mine running - no problems so far

Regards

Graham

NODE: EumetcastPC_HVS-2 
RECV: Quad Inverto Black Ultra TBS-6909X V2 PCIe card TelliCast client(s) 2.14.7           <--- I like it flush :-)
DESC: Receives BASIC (MODCODS 8PSK 3/5) + HVS-1 + HVS-2 ... using 3 demodulators 
-------------------------------------------------------------------------------- 

Graham,

AFAIK you have a TBS-6909X V2 {:-)>>. GNU/Linux works well ...

Regards,
Ernst

P.S. Maybe you can give a couple of other people a hand to use TelliCastStatsCollector. It's very easy now as you don't have to mess with additional tools and libs ...


David J Taylor GM8ARV 🏴󠁧󠁢󠁳󠁣󠁴󠁿 🇪🇺
 

On 21/10/2021 18:11, Alan Curnow wrote:
Ernst,

Local time here is always 1 hour ahead or equal to UTC so it hopefully that
shouldn't be a problem. The receiving PC doesn't do anything else but Tellicast
so I could change it to UTC I suppose.

Does setting a scheduled task to run indefinitely from a specific time continue
after a computer restart? Presumably it does if that's what David does.
[]
Cheers,
Alan
Alan,

Yes, running indefinitely starts correctly after a reboot (at least assuming
the start date is /before/ the reboot.

Internally all Windows systems run in UTC, and I would hope that Ernst's
scripts do as well. My software for the statistics include a program returning
the UTC date and time. My PCs also run in local time - UTC+1 at the moment.
My phone runs in local time, but I keep my cameras on UTC.

Cheers,
David
--
SatSignal Software - Quality software for you
Web: https://www.satsignal.eu
Email: david-taylor@blueyonder.co.uk
Twitter: @gm8arv


Alan Curnow
 

Thanks David,

I've changed this and the ramdisk trimmer to run indefinite from a time now so all should be well.

I've also just swapped my USB TBS5927 to run off a USB-C port as it was previously in a USB3 port alongside the SR-1 which connects via a Gigabit Ethernet to USB3 adapter. The USB3 ports probably share a controller so there may have been some interaction causing lost packets, which have increased quite a bit on Basic since installing the TBS. Time will tell.

Cheers,
Alan


David J Taylor GM8ARV 🏴󠁧󠁢󠁳󠁣󠁴󠁿 🇪🇺
 

On 21/10/2021 19:37, Alan Curnow wrote:
Thanks David,

I've changed this and the ramdisk trimmer to run indefinite from a time now so
all should be well.

I've also just swapped my USB TBS5927 to run off a USB-C port as it was
previously in a USB3 port alongside the SR-1 which connects via a Gigabit
Ethernet to USB3 adapter. The USB3 ports probably share a controller so there
may have been some interaction causing lost packets, which have increased quite
a bit on Basic since installing the TBS. Time will tell.

Cheers,
Alan
Alan,

I know Ian Deans reported problems with the TBS5927. I found that there were
inexplicable losses - either missed and recovered (which hide actual missed and
recovered) or lost, which lead to data loss.

I eventually swapped the TBS5927 for a direct network feed off the Ayecka SR1.
No losses, or missed & recovered now.

In a different context I do recall adding a separate USB 3 PCIe card to get a
higher sustained data rate throughput - at least that has its own controller.
Renesas, IIRC.

Cheers,
David
--
SatSignal Software - Quality software for you
Web: https://www.satsignal.eu
Email: david-taylor@blueyonder.co.uk
Twitter: @gm8arv


Ernst Lobsiger
 

On Fri, Oct 22, 2021 at 02:25 AM, David J Taylor GM8ARV 🏴󠁧󠁢󠁳󠁣󠁴󠁿 🇪🇺 wrote:
I know Ian Deans reported problems with the TBS5927. I found that there were
inexplicable losses - either missed and recovered (which hide actual missed and
recovered) or lost, which lead to data loss.
David and Alan,

as usual it's either hardware (HW) or software (SW).

HW may mean the USB box itself or the USB system HW of the PC.
SW may mean the driver or in case of Windows 10 IP-Tool or BDADataEx.

I doubt that the box has a HW problem. If many users have problems with the TBS5927 while they say the TBS5925 runs just fine this points to SW driver problems.
As a similar issue with PCIe cards Graham Woolf could not get the TBS6909X running under Windows 10 while the card is an outstanding receiver under GNU/Linux.
To see whether the TBS5927 performs under GNU/Linux much better than under Windows 10 we must find the right person (maybe Youssef Bennouna can comment?).

I have no idea whether there are tools under Windows 10 to analyze the data stream e.g. for Continuity Counter Errors of the MPE packets (easy under GNU/Linux)
Once the TBS5927 problem is identified with a high probability as a Windows 10 driver issue someone should go for it and do whatever is possible to get that fixed.

Cheers,
Ernst


Graham Woolf
 

Hi Ernst

Here are my latest files

Personally I would like to replace the current delimiter with a comma rather than the apostrophe like here

183,31,'404,454

rather than

183'312'404'454

Can you point me at the line to change in the file

Many thanks

Graham


Ernst Lobsiger
 

On Fri, Oct 22, 2021 at 05:28 AM, Graham Woolf wrote:
Personally I would like to replace the current delimiter with a comma rather than the apostrophe like here

183,31,'404,454
Graham,

I doubt you really want the above? Assuming this is just one of your typos: Everything is clearly documented in the parameter section of the script!

<CITE>
:: Numeric delimiter character: Set this ' or . or , or whatever you think looks good
set Ndel='
</CITE>

As usual you have to RTFM and please also read this:
<CITE>
:: Double check that you set variables without leading or trailing blanks!
</CITE>

Cheers,
Ernst