Re: Gaps in AVHRR images yesterday.


geojohnt@...
 

David,

Thanks for your further comments.

I run MSG-4, IODC and RSS MDM's but in reduced channels.
The above MSG-4 MDM FSD FY-2H, FY-2G both 6 channels selected - GOMS all channels selected but it  seems to have disappeared.
AVHRR Manager NOAA 19, Metop A, B and C.
MSG Animator Europe cropped IR - Today's date only.
Himawari - VIS, IR1, IR2, IR3, IR4.
GOES 16 and GOES 17 both full 16 channels appear selected - can these be trimmed?
Metop Manager x 2 for B and C.

I 'take' Modis data but don't have the programme open.

And in MSG-4 MDM Other - 1, ATOVS & sounder data Manage Ch. 1 data is ticked and all boxes below ticked ATOVS retention 1.
  
'Others-2' only Modis L1/L2 data Manage MODIS L1/L2 and Copy M02 sensor files is ticked - Data retention 1.
'Others-2' also only Ch. 12 data all three boxes ticked with Fire retention 1.

Sorry everyone if this is all a bit tedious.
Hopefully for some users it might be helpful?

One thing you mentioned that could be relevant - Anti-virus software.

As you know I use the 'dreaded' BullGuard software.
This computer running 24/7 is used for EUMETCast only but has an open Broadband Internet connection 24/7 for Windows and SatSignal updates.
EUMETCast data is (appears to be) going through BullGuard, as well.

Regards,
John.
 
+++++++++++++++++++++++++

John,

I am not completely happy with SSDs for EUMETCast data because they have a 
limited number of write cycles, a limited total data throughput, resulting 
in a limited lifetime.  Perhaps with today's SSDs these limits are such that 
it's an acceptable choice,and if the user is aware of that, fine.  Remember 
that there's not only receiving the data, but there might be processing as 
well.  For HVS-2 it's something like 450 GB per day from the source.

HDD - slow
SSD - fast
RAMdisk - much faster

But RAMdisk is limited in capacity and may need more management.  Some of my 
systems are now using RAMdisk (1-2 GB for BAS, 10 GB for HVS-2 
[receive-only]) with Robocopy moving the data from the RAMdisk onto HDD once 
a minute, and with TrimTree running regularly to remove any lingering "tmp" 
files.  One HVS-1 system writes directly to a separate HD, One BAS system 
works well without any RAMdisk.

It's other things which can help too - preventing any unwanted disk access 
from e.g. indexing and anti-virus.  Minimising the data you take.

One thing which I'm increasingly finding critical is to maximise the number 
of data paths through the system.  Perhaps add a second HDD controller 
rather than just another SATA connection.  I'm currently thinking about 
adding a second USB internal hub for the TBS5925 as the PC has a single 
master hub.  There's an element of "try it and see" about this level of 
changes.

Cheers,

David

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

++++++++++++++++++++++++++++++++++

-----Original Message-----
From: David J Taylor GM8ARV 🏴󠁧󠁢󠁳󠁣󠁴󠁿 🇪🇺 via groups.io <david-taylor@...>
To: MSG-1@groups.io
Sent: Thu, 24 Sep 2020 16:00
Subject: Re: [MSG-1] Gaps in AVHRR images yesterday.

David,

Thanks for pointing me (again) to TD 15.
It really is now quite comprehensive.

And I still have my 999 MB RAMdisk unused which I was going to (will)
experiment with.
As you know Ops want me to do a couple of tests on my current setup which
I'll do hopefully under clear sky conditions - when they return.

I've also thought of having a second HDD for HVS-1 alone.

[I have an SSD for 'Windows.']

TD 15 also recommends RAMdisk - and interestingly - SSD's which you used not
to favour(?) as it points out HDD's are 'slow' in these days of high data
level I/O's..
And, Ernst has pointed out that the new generation of HDD's are 'slower' and
best avoided.

So I have a couple of updates to consider to 'speed things up?'

Regards,
John
=====================================






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