|
Grex > Helpers > #137: Grex System Announcements - Winter 2004/2005 |  |
|
| Author |
Message |
| 25 new of 219 responses total. |
naftee
|
|
response 139 of 219:
| 
Jan 13 05:12 UTC 2005 |
re 137 Yeah, whoops. Silly me for thinking all URLs were not case-sensitive.
|
juicy
|
|
response 140 of 219:
|
Jan 13 06:02 UTC 2005 |
nope, just the domain.
|
gelinas
|
|
response 141 of 219:
|
Jan 21 23:56 UTC 2005 |
Grex will be unavailable from 12:00 to approximately 15:00 tomorrow,
22 January 2005, while it is moved from the Pumpkin to provide.net.
The old machine will move back to its old IP address, but logins (and mail)
will be disabled; it is moving back to provide DNS service for our domain
until our domain registration is updated.
|
mfp
|
|
response 142 of 219:
|
Jan 22 01:02 UTC 2005 |
pwho
bbs
r
r
w
pwho
bbs
bbs
pwho
|
keesan
|
|
response 143 of 219:
|
Jan 22 04:39 UTC 2005 |
Does this mean grex will be available via the website using the old machine?
|
gelinas
|
|
response 144 of 219:
|
Jan 22 05:04 UTC 2005 |
I don't quite see how you reached that conclusion, but no: www.cyberspace is
an alias for grex.cyberspace, and so will move along with grex. I mentioned
logins because _some_people will probably try to connect by IP address
instead of domain name.
|
keesan
|
|
response 145 of 219:
|
Jan 22 05:18 UTC 2005 |
Why is it not possible to hook up the old grex to the same ISP connection and
DSL modem to check if the problem is in the computer or in something else
(such as the DSL modem or the connection to the ISP)? Would this interfere
with using it as a DNS server? And why do we need a DNS server if grex is
not working?
|
gelinas
|
|
response 146 of 219:
|
Jan 22 06:20 UTC 2005 |
oldGrex _is_ on the same connection as the current grex. I log into it
regularly. However, since I'm the only user, there is not much going on.
There are other machines on the network as well, which other staff members
use regularly. They have mentioned seeing similar behaviour there.
We need a DNS server so that people can find our machine by its name. This
is one of the functions of the main machine. However, DNS "bootstraps" from
an IP address. Right now, that IP address is 216.93.104.34. Until I can get
the new IP address registered, and so get the 'bootstrap' fixed, we have to
keep a DNS server running on the old address.
|
keesan
|
|
response 147 of 219:
|
Jan 22 11:54 UTC 2005 |
Similar or ientical behavior? Could we have two different problems at once?
The old problem was 2 minute freezes every few minutes. This is 10 sec
freezes every couple seconds. The old problem occurred during telnet, not
dialup. What is the motehrboard maker and model? Please email
keesan@freeshell.org
|
gelinas
|
|
response 148 of 219:
|
Jan 22 21:11 UTC 2005 |
ASUS, and I looked at the capacitors; they are fine.
Grex is now in the new location.
It will take a bit longer to get the dial-ins re-directed.
|
naftee
|
|
response 149 of 219:
|
Jan 23 02:17 UTC 2005 |
gelinas, why don't you refer ms. keesan to a text on networking, and save
yourself some time and energy? I hate to see you stressing out.
|
gull
|
|
response 150 of 219:
|
Jan 23 22:36 UTC 2005 |
Everything is nice and fast now. Good work.
I was a little confused yesterday when I connected via dialin and got
something like this after logging in:
This is _NOT_ Grex. Use the domain name, not the IP address.
Dialin Users
NO CARRIER
|
keesan
|
|
response 151 of 219:
|
Jan 23 23:23 UTC 2005 |
Ssh connection is nearly instant. Telnet not working, but it worked this
morning, very slowly.
|
naftee
|
|
response 152 of 219:
|
Jan 24 02:16 UTC 2005 |
Just re-booted.
|
keesan
|
|
response 153 of 219:
|
Jan 24 04:10 UTC 2005 |
Telnet and pine are working again but I have not received any mail for a few
hours except something from grex. I would have expected 10 spams in this
period.
|
gelinas
|
|
response 154 of 219:
|
Jan 25 18:44 UTC 2005 |
Grex panicked and did not reboot itself Monday evening.
|
gelinas
|
|
response 155 of 219:
|
Jan 25 18:49 UTC 2005 |
I think I fixed the terminal server to connect to the new machine last night,
so dial-in should be working again.
I also think I fixed the web proxy to accept requests from the new machine.
|
scholar
|
|
response 156 of 219:
|
Jan 25 19:30 UTC 2005 |
Thanks, Joe!
|
keesan
|
|
response 157 of 219:
|
Jan 26 03:15 UTC 2005 |
What does 'grex panicked' mean in more technical terms? Is this some odd
hardware problem or a response to something done to the software?
|
janc
|
|
response 158 of 219:
|
Jan 26 03:43 UTC 2005 |
Software.
|
keesan
|
|
response 159 of 219:
|
Jan 26 03:46 UTC 2005 |
I get a kernel panic when I tell it there is more memory than actual, or to
put root on some nonexistent device. Did you do something equally clever?
|
janc
|
|
response 160 of 219:
|
Jan 26 03:58 UTC 2005 |
I installed a slightly newer version of backtalk and fronttalk. Backtalk has
suffered quite a few internal reorganizations aimed it making various things
I want to do in the future easier to do. There's always a risk of introducing
new bugs when I do this, but that's life. The major bug fix in Backtalk is
to the date parsing. "read since" type commands should work much better.
Fronttalk has had more noticable changes. First, it now has command line
editting (via a package known to Unix geeks as "readline") similar to what
you see in tcsh and other modern shells. That means you can use the arrow
key to move back into the command you are typing and edit it, or use the up
arrow key to move back in the command history and reissue previous commands.
The "read since" command was working badly. It should now work much better.
The date syntax understood by Backtalk isn't exactly the same as Picospan or
Yapp - it actually accepts a much wider range of date syntaxes. Doing
"help date" in Fronttalk will tell you more than any sane person wants to
know about Fronttalk date formats. Backtalk now implements the Picospan
"date" and "cdate" commands, but they are pretty useless, so don't worry
about it.
Control-Z generally should work better now in Fronttalk - things should
work smoothly when you suspend/restart the program. Actually, part of this
was a BSD portability fix in Gate, so it may fix sometime problems for
Picospan users too.
I would encourage regular users of "bbs" (Picospan) to try out Fronttalk.
You run Fronttalk with the "ft" command. The plan is that Fronttalk will
eventually replace Picospan on Grex. Trying it out now to make sure it
behaves well for you would be wise.
|
janc
|
|
response 161 of 219:
|
Jan 26 03:59 UTC 2005 |
Re 159: No, this is probably an OpenBSD bug. There is a theory that it is
due to using soft updates, a feature that may not be fully mature in OpenBSD
version 3.5. We don't actually know what caused the last crash though.
Things are being tried.
|
twenex
|
|
response 162 of 219:
|
Jan 26 12:07 UTC 2005 |
A kernel panic is the UNIX version of a Windows "Blue Screen of Death" -
except that the last time it happened every day was probably in 1971 or so.
For the record, Windows isn't the worst - IBM's ill-fated TimeSharing System,
TSS, took over 10 minutes to boot up, but mean time to failure was less than
that. D'Oh!
|
keesan
|
|
response 163 of 219:
|
Jan 26 13:55 UTC 2005 |
I am trying out fronttalk now and when I hit the left arrow I get [D and a
beep. Am I supposed to be able to back up? The up arrow give me [A and a
beep. I have an 84-key keyboard and am using linux. Down arrow [B. Right
arrow [C. Backspace works as expected.
|