Re: TC V2.14.4 and RAMDisk?

geojohnt@...
 

David,


Thanks for your useful comments.
I've just got in from being out all day and need time to digest your information.

Actually, I defragged my 1.81 TB data hard disk last night - it said it was 75% fragmented.
OK, I have been deleting some GB's of accumulated non wanted 'non imaging' files recently but defragged it a couple of weeks ago.
Anyway, HRPT Reader and GeoSatSignal now work much faster.
And I don't, apparently, get any missing segments/chunks.

So, 'if it ain't broke .......'

I was just surprised that EUMETSAT appeared to be suggesting a RAMDisk as well as temp folders with the 'new' TC.
But I guess if you are taking everything, then two paths (or more) are better than one.

Goodness, an 8 GB RAMDisk?!

Regards,
John.

+++++++++++++++++++ 

In a message dated 07/07/2019 17:58:59 GMT Standard Time, gm8arv@... writes:

Hello Group,

I wonder if there are users in the group with a one PC receiving/processing
system, receiving all BS and HVS-1 imaging and using a RAMDisk.

As many will know I've at last moved to TC V2.14.4 for BS and for the
reception of HVS-1 channels with 'varying degrees of success.'

I'm running:
MDM - MSG-4, IODC, RSS GOES-16, GOES-17, Himawari - and Animator for RSS.
AVHRR Manager, 2 X Metop Manager.

All 24/7.

I have good BS SNR, and 'more than adequate' on HVS-1 from a 1 m dish and
Inverto Black Ultra LNB

Now, my single PC system computer - i5-4690K @ 3.5 GHz, 16 GB RAM, 64 bit
Windows 10 - appears to have slowed down since having all the extra
programmes running in regards of opening and loading data for HRPT Reader
and GeoSatSignal.
Not a problem but one now has to wait a little longer.

I noticed since upgrading to the 'latest' TC, that my previous RAMDisk of
997 MB is not being used.
OK, I know that TC V2.14. xxx is designed to do the job of RAMDisk in temp
folders, and I see it does just that.

However, I had a look at the 'latest' EUM TD 15 yesterday and see that it
has quite a lot of information on configuration and using a RAMDisk.
Not only that, suggesting(?) the use of SSD's against HDD's, for their
speed.
Pages 34 - 35.

I know that SSD's are not in favour by some(?) but I wonder if anyone has
any comments/experience on 'still' using a RAMDisk and using SSD's in the
data path?

Regards,
John Tellick.
=================================

John,


Others have used SSDs successfully, although I haven't here.  You may like
to check the life expectancy given your expected data flow - I recall one
system reported here had three SSDs, one for each service.

My own systems are:

PC Alta: receive and process a selected set of the BAS data.  1000 MB
RAMdisk.  No HVS processing or reception.  240 GB SSD system disk, two HDs.
  https://www.satsignal.eu/mrtg/alta-ramdisk.html

PC Stamsund: receive and process selected BAS with 1300 MB RAMdisk.  Process
HVS-1 data received over the network from PC Lund.  Receive but not process
HVS-2 data, it's deleted soon after reception.  Two HDs.
  https://www.satsignal.eu/mrtg/stamsund-ramdisk.html

PC Kiruna: process selected BAS data received over the network from PC
Harstad.  Receive all HVS-1 data direct to separate HD (no RAMdisk) with
selected image data processed.  256 GB SSD, three HDs.  No RAMdisk

PC Harstad: receive-only selected BAS data, temp & received on 4000 MB
RAMdisk, moved to HD with Robocopy.  Data fed to PC Kiruna.  Single 1 TB HD.
  https://www.satsignal.eu/mrtg/harstad-ramdisk.html

PC Lund: receive all HVS-1 and HVS-2 data.  HVS-1 data temp and received on
10 GB RAMdisk, moved as soon as possible to 4 TB HD using Robocopy, where it
is served over the network to PC Stamsund.  HVS-2 data temp & received on
the same 10 GB RAMdisk, but deleted soon after reception.

  https://www.satsignal.eu/mrtg/lund-ramdisk-trees.html

They "just grew" like this!  There's quite a lot of data I don't use, but I
want to keep receiving all the HVS-1 and HVS-2 data so that I can compare
with Arne.  We both submit automated daily reports to EUMETSAT.  I have
TrimTree jobs doing this, which also remove unneeded processed data (e.g.
much of the sounder data).  It just grew like this.

All of the systems are capable of loss-free reception when conditions are
right.  Instantaneous signal drop-outs are better handled by the TBS
systems, possibly due to a faster AGC response (I've been unable to confirm
this), the system most susceptible to missed data due to disk activity is
Stamsund, without a separate system disk.  All systems running latest
Windows-10 (1903) and latest TelliCast.

Arne's system does what you want (all three services), with an 8 GB RAMdisk,
but he uses a much more complicated script.  My Robocopy/TrimTree method
does the same job, as far as I understand.

Your turn, you choose!

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




Join MSG-1@groups.io to automatically receive all group messages.