Re: Timezones, .55 server down, graphing
Hi, Andrew -- I'll fix these when I get back from Toronto (for AGR'04) next
week. I got lots of good feedback from the folks there, and so have other
things to fix, too. ;-)
I *would* like to look at your graphing code and see what would be involved
with incorporating it into the main Beacon release, please.
The 600 second mark in the history file code is when the current history
file is moved to the prevhistory file, and the current history file is
started over. That shift occurs during the WEBREFRESH invterval update
(currently 60 seconds) that comes after the 600 seconds has expired.
Thanks for your feedback -- It's been quite helpful.
Mitch
At 10:06 AM 6/12/2004 -0700, you wrote:
>
> On the central server page, it says the time is in CST. It's actually in
> CDT (I presume it's in "central" and changes with the seasons)
>
> The 9.55 server at http://beacon.ncsa.uiuc.edu/ seems to have
> died at Sat Jun 12 0:04:14 2004 CST without changing to the
> "stopped" page.
>
> The time in the history file is C?T I guess so my graph is in Central,
> not Pacific as I'd put initially. Duh; should have caught that..
>
> (I guess my mail did go through; I was looking in the wrong month..)
>
> Yes, my code's available, but not nicely packaged. Please ask me directly.
>
> The Y axis is in arbitrary units. It's possible to plot jitter and loss
> on the same page (a bit silly, perhaps). There's a mark at top right that
> says what the different traces are. I can add a Y label when there's only
> one trace - probably a good idea
>
> --
> Andrew Daviel, TRIUMF, Canada
> Tel. +1 (604) 222-7376
> security@triumf.ca
>
>
--
Mitch Kutzko | mitch@dast.nlanr.net | mitch@ncsa.uiuc.edu | 217-333-1199
http://hobbes.ncsa.uiuc.edu/