TO: All GCN Sites
RE: Changing the GCN Machine
DT: 10 Mar 03
This is the third notice (the first notice with all the details is below).
Everything is still on track for the switch-over on 11 Mar 03.
There is one new development related to email sites. I had originally thought
that only the socket sites with firewalls would have to make a change,
but I have since learned that at least one email site also uses the firewall
method to protect against unwanted incoming email. So EMAIL sites
should also check with their network administrators (a) to see if they
have any such filtering at their end, and (b) to get that firewall updated
with the new machine number (see 1st notice below for the number).
Sincerely,
Scott Barthelmy
GCN Ops
//////////////////////////////// First Noteice /////////////////////////////
TO: All GCN Sites
RE: Changing the GCN Machine
DT: 21 Feb 03
PUNCHLINE and ACTION ITEMS:
1) Socket sites that have firewalls will need to update their firewalls.
2) All other sites will not need to do anything.
Read on for the details.
THE PHYSICAL CHANGE:
The machine and network connection for the GCN system will be changed
to a new machine and a new network connection. The change is scheduled
for Mar 11, 2003.
NO APPARANT CHANGES FOR THE EMAIL-BASED USERS:
The change will be totally transparent to the email/pager/cellphone sites.
The machine name will stay the same ("capella"), and the account name ("vxw")
will stay the same. This means any site doing 'filtering' on their
incoming email will not need to change any of their filters. The From-line
will stay exactly the same. (And of course, all the formats of the emails
will stay the exactly the same.)
SOME CHANGES REQUIRED FOR SOME SOCKET USERS:
Some of the socket sites will require a change at your end.
Only those socket sites that have firewalls at your end will need to change
the "hole" in your firewall for the new IP Number for the new "capella"
machine. Unfortunately, the Goddard network will not allow the new capella
machine to have the same IP Number as the old capella machine, hence the
change in the IP Number (new number is: 128.183.16.187). However,
all the port number assignments to each socket site will stay exactly the same.
(And of course, the content and formats of all the sockets packets will stay
exactly the same.) It is recommended that the those sites with firewalls
leave the old IP Number entry (128.183.219.100) in their firewall tables
in case there are problems with the switchover and I have to go back
to the old system -- two weeks is a reasonable duration.
It is also recommended that the firewall-protected sites add a second new
IP number into the list. The second new number is 128.183.16.188.
This second machine is a backup machine in case the primary machine
would ever fail. It is a completely redundant system. Hopefully, I will
never need to switch to it, but in case there is a problem sometime in the future,
then this backup machine is already in place in your firewall to accept
incoming GCN socket traffic.
WHY THE CHANGE:
The current machine is getting old; there is an increasing possibility
of a hardware failure. Also, new government security regulations
can not be met with the current operating system. And day-to-day operations
will be easier if the GCN machine is closer. It is currently 4 buildings
across the Goddard campus. Now it will be inside the same building
that my office and the system administrators are located.
SWITCH-OVER OUTAGE DATE and DURATION:
The switch-over is scheduled to start ~15:00 UT on Tuesday 11 Mar 2003.
It should take approximately 3 hours -- GCN will be off the air for 3 hours
(until 18:00 UT). Most of that time has to do with getting the GSFC name server
updated with the new information about the new machine with the same old name.
IN CASE THERE ARE PROBLEMS:
The old GCN system will be maintained in an off-the-internet configuration.
So if there are unforeseen problems in the switch-over, we can switch back
the old system within another 1-2 hours (and be back on the air with the
normal GCN operations).
QUESTION/PROBLEMS/COMMENTS:
I believe I have accounted for and protected all aspects of all sites,
but in case I missed something, please tell me as soon as possible.
I suspect special accomodations can be made at this end, but in case not,
I need to know to change the switch-over plan.
It would be helpful to hear back from all sites, even those that have
no problems or no changes to accomodate the new machine. Closing the loop
always makes these kinds of operations go better -- a simple "OK" is
perfectly sufficient. Questions and comments are always welcome.
There will be reminder announcements as we approach the switch-over date.
Sincerely,
Scott Barthelmy
Scott Barthelmy NASA-GSFC, Code 661, Greenbelt, MD 20771
PHONE: 301-286-3106 (office)
FAX: 301-286-1684 (1st choice, -1682 2nd choice)
EMAIL: scott@lheamail.gsfc.nasa.gov
PAGER: 3013134808@my2way.com (by email, ~200 characters/message)
WEB: http://gcn.gsfc.nasa.gov/gcn
////////////////////// Your site's current configuration //////////////////////
Site_name: VSNET
Lon,Lat= 135.75 35.00
Distribution_method: EMAIL
Address: vsnet-grb@kusastro.kyoto-u.ac.jp
Filter: ALL
Error_limit= 360.100 [deg, diameter]
Delta_T_limit= 999.900 [hr]
Sources & Notice Types/Subtypes:
RXTE-PCA_Alert_wont: Enabled
RXTE-PCA_Alert_will: Enabled
RXTE-PCA_nosaw: Enabled
RXTE-PCA_saw: Enabled
RXTE-ASM: Enabled
IPN_POS: Enabled
ALEXIS: Enabled
RXTE-ASM_TRANS: Enabled
HETE_S/C_ALERT: Enabled
HETE_S/C_UPDATE: Enabled
HETE_S/C_LAST: Enabled
HETE_GND_ANA: Enabled
HETE_Test_notices: Disabled
Test_type2_notices: Disabled
INTEGRAL_Test_notices: Disabled
INTEGRAL_POINTDIR: Disabled
INTEGRAL_SPIACS: Enabled
INTEGRAL_WAKEUP: Enabled
INTEGRAL_REFINED: Enabled
INTEGRAL_OFFLINE: Enabled
When requesting changes to your configuration,
it helps if your refer to your site_name.

Return to the Powerful Daisaku Nogami
vsnet-adm@kusastro.kyoto-u.ac.jp