[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

19991215: reply from David Knight re: NLDN testing

------- Forwarded Message

>To: address@hidden (Unidata Support)
>cc: address@hidden
>From: "David J. Knight" <address@hidden>
>Subject: Re: 19991214: NLDN Y2K IDD header discussion (cont.)
>Organization: .
>Keywords: 199912152233.PAA25759

Tom,

Sounds like a plan.

We cannot change the date on aspen, but, it is not our operational
machine for any IDD product so can be used to test the fixed up nldn
code. Feel free to run an ldm on it as well if you like.

I'll start looking for a suitable machine for further testing. Pete
perhaps you have some suggestions??? cessna comes to mind but I hate to
saddle them with that old dog! maybe beech for a few days? Else I'll
just pull one out of the user area - spruce?  We'll probably have to
remove that machine from our nis+ database for awhile, and install a
local account nldntest (or something) because I do not want our
NFS/NIS+ server seeing activity from the future - also make sure we
disable automatic time syncing!

This is beyond my normal area of experience.  Perhaps there is really
no problem with having the NFS server seeing file creations in the
future?  Would sudo date serve your purposes?  I'm wide open to
suggestions on the best way to do this!

For now much can be gained just by making sure the revised code works
properly on aspen. Even though I have little C experience, I'll be
curious to find out what had to be changed, and how it went...  (Sandy
probably saw some Mitch B. code from when he was here).

It might be a few days till we can identify and configure a test
machine for you. This is complicated by the fact that the NLDN
satellite receiver upgrade that was supposed to happen yesterday has
had to be rescheduled. Also the end of semester is near, so demand
tends to be high.

David

> >From: Unidata User Support <address@hidden>
> >Organization: Unidata
> >Keywords: NLDN IDD pqact.conf WMO header
> 
> David,
> 
> I just talked with Sandy about the progress he has made with the NLDN
> programs he FTPed from aspen; things look pretty good.
> 
> I understand from your last eamil that 'aspen' is not a machine on
> which we can do NLDN Y2K stress testing, but you did indicate that you
> might be able to find a machine that could be used.  Since we feel that
> doing a reasonable test does involve changing the clock on the intest
> machine to simulate the transition into the year 2000, I am asking you
> to please find a machine on which we can do this.  Also, we will need
> root access to this machine so we can change the time at will.
> 
> In the interim, I need to make sure that 'aspen' is not your
> operational NLDN machine so we can:
> 
> o rebuild the various executables on it
> o see if they work on a live feed for an extended period of time
> 
> Assuming that the above is successful, we need to:
> 
> o build the various routines on a test machine on which we change the time
> o get things running as above
> o change time and watch for problems
> 
> I will probably mail the NLDN community that they need to modify their
> NLDN pattern(s) in their LDM's pqact.conf file tomorrow.  I will advise
> them to change pattern to be able to match on either two or four digit
> years so that we have flexibility in the futue to change the product
> file name from YYJJJHHMbMe to CCYYJJJHHMbMe.
> 
> Any help you can provide in getting us a good test machine will be
> greatly appreciated.
> 
> Tom,


------- End of Forwarded Message