222 new of 237 responses total.
Could be mh is misconfigured too. There are lots of ways it could be trying to send mail, and some of them probably will lose.
This response has been erased.
No, but I hadn't used MH for outgoing mail before. I don't usually send _any_ mail from grex. So until now, Pine has been sufficient for the little bit of test-mail I've sent.
This response has been erased.
There's a new RISKS digest out, but I cannot get to ftp.sri.com to get it here.
I got it in email, but I already deleted it. Sorry.
I'm having a bit of a backtalk issue. When using the "view responses" textbox, and pressing enter (while in the textbox), rather than showing me those responses, it takes me to the "Set Private Item Title" screen
Which browser? Browser behavior varies on this one.
I noticed an extreme slogginess in grex response this morning. When I checked top, it showed some pretty hungry sendmail processes. The load was over 2 which probably isn't horrid, but if those sendmail documents were of any volume, the pipeline was probably stressed. If grex is supposed to handle large volumes of mail, then ignore this report.
re #23 - Opera
OK, I thought it might be Opera. Valerie had the same problem with Opera, but I haven't seen it with any other browser. We tracked it down a bit further and, though I don't remember the details, what Opera was doing was simply weird. I think Valerie submitted it to them as a bug report.
The chantab file /var/spool/party/chantab is corrupted. There are several places where there are partial lines followed by a series of blank lines and then additional partial lines, which do not match up to the previous partial lines. (It's not a case of several carriage returns appearing in the middle of a line.) Can a clean copy be pulled from a backup, or should a partyadm just go through and clean up as best we can?
Received a few telegrams this afternoon from people asking that an item be posted about /c being full or nearly full.
Why was grex not answering the phone for the past 24 hours or so?
It was down. I rebooted it. I think the UPS isn't doing the right thing, but I don't know.
Can you put that in the motd please?
Re #30: The modems are only down when the power is down, so that would tend to implicate the UPS.
Why did I get 86 brandnew items when I just entered Agora?
Sounds like your participation file got trashed as a side effect of /c (where your home directory is) filling up.
With Scott and Valerie mostly out of the disk police business, staff isn't as much on top of disk space as we have been in the past. I did a quick pass through things deleting some huge disk hogs. That should at least temporarily keep things going.
Someone to whom I sent a binary file as an attachment reports that it arrived (in 7-bit format) as part of the message body. This was a translation and it has wasted quite a bit of time to figure out what went wrong and resend. What might be happening? At other times the attachments have not arrived at all.
Did you set the file transfer to binary?
Yes, I always send kermit -ir. It was coded base 64 and remained part of the message body. Maybe his mail program goofed?
The modem on -3596 appears to be hosed. It connects, but it will not transfer any data (it went down abruptly this morning and is still not working).
The modem on 761-3596 is STILL hung; it connects but never issues the "connection established" message. The other modems are not equivalent; the one on 761-3000 transfers data at less than half speed.
Attempting to connect via ssh, I received an error I'd not seen before: 11:38pm jis8 $ ssh dcat@grex.org Warning: Server lies about size of server public key: actual size is 767 bits vs. announced 768. Warning: This may be due to an old implementation of ssh. The authenticity of host 'grex.org (216.93.104.34)' can't be established. RSA1 key fingerprint is 74:4c:1e:86:04:2a:ac:ab:c2:cd:32:ff:19:76:80:fc. Are you sure you want to continue connecting (yes/no)? yes Warning: Permanently added 'grex.org,216.93.104.34' (RSA1) to the list of known hosts. dcat@grex.org's password: File Size Limit Exceeded And with that I was dumped back to the system from which I was trying to connect. On the second try I was able to connect normally, and was nowhere near any "File Size Limit" -- I've little in $HOME, and no mail stored : grex:11:39PM% du -hks ~ 270 /c/d/c/dcat grex:11:39PM% mailx No mail for dcat Is this another problem with the sshd, like the previously-reported 'size of public key' error, or something else?
I couldn't dial in to Grex just now. The modem on 3000 answered, but never connected me to grex or a login request.
Grex was down some today while STeve worked on the UPS.
Grex is not answering the phone. Even after Steve worked on the UPS this Saturday.
STeve didn't finish the UPS. He needs 2 more batteries.
I get the "Server lies" bit off and on all the time. It's harmless, an artifact of Grex's old version of SSH. The fingerprint issue should only happen the very first time you connect to Grex from a new machine.
Are the batteries series or parallel connected?
resp:46 -- as I noted, I was aware of the 'size of public key' error; it was the "File Size Limit Exceeded" error I didn't understand, and still don't.
Grex has been up (telnet, web, etc. okay) but not serving the dial-in lines (CONNECT string is all you get) several times in the past day or so.
As I write this, Grex has had a partial or complete modem outage for ~24 hours, with no end (or diagnosis) in sight. The modems answer, but "you can't get there from here". Why the people running this show can't turn the modems off if they aren't going to be allowed to work, I don't know. Is this malign neglect, or merely gross negligence? Thanks for cutting off (and disenfranchising) paying members! </sarcasm>
Gee, maybe we should make the staff take an unpaid leave of absence as a punishment.
Fire those incompitent bastards! I suspect STeve turned the modems off when he was working on the UPS; it's what we used to do whenever Grex was down. These days it's possible to leave a message on the terminal server, but I'm probably the only one who knows how to do that. I'll send mail about it.
I was at the pumpkin doing a reboot today, but since I didn't know the modems were broken, I didn't try to fix them. I don't think I even have a computer with a modem hooked up to it (wait - there might be one in my laptop). I have no simple way to tell if Grex's modems are working. I did power cycle the modems while I was rebooting Grex. I probably should have power cycled the terminal server just on general principle. If it was booted up before gryps, then it probably wouldn't have been able to get it's updated firmware from the tftp server. That might explain this behavior. I suppose I could take an hour to go back to the pumpkin and do that, but I'm far from enthusiastic. I won't even be able to tell if it worked.
STeve ewailed that it was his fault, and that he'd go fix it.
OK, so my iBook does have a modem. Does anyone know how to get a communications program for it? I spent half an hour searching the web for information, and didn't find anything useful.
What's the OS, Jan? If MacOS X, . . . oh. Yeah. We don't do PPP, do we? :( Used to be, I'd recommend VersaTerm, but I don't know what to use nowadays.
I used to use ZTerm when I had a Mac. I don't know if it's OS X compatible.
Ah, I found ZTerm for OS X. Seems to work. Now I'm sure I have a modem cable around here someplace....
Any modular telephone cord will work. Just unplug the regular telephone, if you like. ;)
I'm sure Jam knows that.
It wouldn't shock me if Apple used some super-special connector for their modems. It'd be like them.
Well, it's Monday morning at 10:00 and *still* I get "modems connect but no Grex." I'm telnetting in from M-Net to report this.
STeve drove over to the Pumpkin to check them out last night. They were up and running.
Did he dial into one? The modems have been up and running all the time. But people connecting to them get the terminal server, but not Grex (or so I'm told - still haven't tried it myself). I think it's the terminal server that needs a reboot, not the modems.
Still cannot dial in.
All that i get is the "CONNECT..." line when i dial in on -3000. Sounds like the terminal server (or something further in) isn't up to speed.
Grex seems to have been off the net between 1am and 11am. Grex didn't actually crash. Just the DSL connection died. I power-cycled the link modem and the connection came up. I also power cycled the terminal server. I'm not sure, but I think someone successfully dialed in.
I'll put a mention in the motd.
I just dialed in.
Thank you janc. We need to get you some helpers with this stuff.
One more week, I hope.
Yup...except we gotta get keys for the new staffers. I've sent mail to baaf about this.
Okay, this is good. The modems are usable again. But why did it take FIVE DAYS for someone to reboot the terminal server after my trouble report last Wednesday?
Staff is having problems at the moment. Nobody is happy about the situation. It's being addressed. Beating people up isn't helping, not really, unless it's to make you feel better. Please give this a little time, maybe a couple weeks, and I suspect Grex will look a little more composed.
This response has been erased.
Sorry for the delay, which was mainly stupidity. For a long time, Grex was pretty stable and Scott was doing the infrequently needed reboots. The UPS developed problems, and hasn't been repaired yet, so the pumpkins dirty power is being fed straight to Grex, and this has caused a series of different problems with Grex. I'm used to completely ignoring modem issues - I habitually ignore them on the presumption that someone else will take care of them. I think the modem problems must have started after STeve shut down Grex to take off the UPS. He forgot to power up gryps before powering up the terminal server, so that the terminal server could download working firmware from gryps. Hey, it's been years since he did it. I wouldn't have remember it either. A few days later I went in to reboot Grex after it crashed. As I rebooted the system, I vaguely recalled people complaining about the modems, so I power-cycled the modems in the vague hope that this would fix whatever ailed them. Only the next day did I actually notice that people were saying that they were connecting to the modem OK, but not getting to Grex. That's a terminal server problem, not a modem problem. I was not eager to go back and reboot the terminal server (each trip to the pumpkin costs about an hour of my time which is none to plentiful these days), especially since I don't even have a way to tell if it worked. I spent some time looking for terminal software for the only computer I own that still has a modem, but before I was finished I heard that STeve had said he would go fix it. I declared myself off the hook. STeve went in...and power cycled the modems. So the fool things still weren't working. This morning, Grex's net connection died. So I went in again. I spent some time power-cycling a gadget that I thought was the DSL modem, but turned out to be just a hub. Eventually I found the DSL modem and power cycled that, which had more pleasing results. While I was there, I power cycled the terminal server (at least I knew what that looked like) though I had no way of telling if it did any good. I got lucky. Russ is right. We did a sucky job on this. STeve and I are both out of practice. There was a time when we did reboots on a daily basis, but it was long ago. Worse, Neither STeve nor I are very appropriate people for this job. We have families and many other obligations, and can not drop what we are doing and scamper off to the pumpkin at the drop of a hat the way we used to. These days we need to work pumpkin trips into our schedules. Maybe adding staff will help with this, though we didn't really recruit new staff members with that criterion in mind. At least it won't hurt. So, anyway, we are in transition. Technical difficulties are occuring. We are aware of them. We are trying to do better. Wish I could promise that we really will do better.
This response has been erased.
That requires further thought, Dan. There are reasons to keep things on separate servers, and there are reasons not to.
My beef with the staff is that nobody was paying attention to problems which had been on record in this item since Wednesday. Suppose that the problem had been reversed, and the DSL modem was down. Suppose further that a staffer had dialed in, logged in, and pronounced Grex fine and healthy and in no further need of attention. Suppose that the lack of Internet connectivity had been noted in this item two days before that.... You get the picture. The fix may be trivial, but it won't happen unless someone pays attention and realizes that it needs to be done.
Yes, Russ. There is no staff person paying any attention to the dial-in lines. We are aware that this is a problem. We apologize for the downtime. We are attempting to rectify the problem. We know you are pissed. A lot more people use the DSL than use the modems. Many of the regular modem users seem to be able to come in over the internet if they need to. The opposite is less true. E-mail service is entirely dependent on the DSL. So the internet connection is higher priority than the modems.
Given how many of the regular modem users are capable of coming here and repeatedly complaining that they're down, I'm not sure why you keep so many modems operating.
Actually there was staff email about the modem problem, starting Sunday I think. Jan, the terminal server tries to get its updated firmware within a few seconds after booting, so just watch its LAN light for activity. Also, if the terminal server does NOT have its updated firmware it won't give a "welcome to Grex" type of message. Useful when complaining to include that message or lack thereofe.
There may well have been staff email about it. I don't get all staff email and I've been ignoring all email about modem for years.
The fact that Russ could get on and gripe about the modems, and then continue with whatever Grex activities he normally does proves that the modems aren't all the critical in contributing to his grex experience. I think it's ludicrous to keep modems and phone lines and deal with the additional expense for the 4 users that use it, especially since nearly all of them can get onto grex through other methods, and it's always nice to save a little money. I understand the sentiment behind it, and the fact that these are old time users, and grex wants to be fair by them, but come on, if they really wanted to be fair to grex, they'd give up on this antiquated method of getting in, and get with the times. Especially since they can get on using other methods and gripe about it. And then bitch about the fact that the staffers pay more attention to the DSL lines more than the modems. (Gee, I wonder why? Is it because without the DSL lines there would be prolly not many more than 4 users on the system? Wouldn't that be such a neat Grex experience?) Staff is doing a great job, considering that they're not paid, and they have other things going on. Give them a break. They've always rallied around at the critical times. I'd like to see what Russ would have to say if they decided they were too busy to get Grex up when it went down.
russ will continue the beatings until morale improves.
I would give up on grex if I had to telnet - it is excruciating. The three people we set up with grex email (one of whom is paying regularly for it) would also be unable to use their email. That makes four of us. Dave Lovelace says he also dials in. That makes five. I would be very surprised if there are not other dial in users. Then there is Jim - 6 (he emails me when the phone is busy). I think we have far too many modems but cutting it to zero would be against the spirit of grex.
If you're dialing in to grex from home, but you CAN telnet in, where are you dialing into to telnet to grex?
Modems constitute free Internet access with extremely cheap hardware. Still part of Grex's mission, I believe. Although we could probably look into cutting a line or two again.
I telnet in and have no problems with slower speed, etc. like Sindi described earlier. Even when I use phone modem as opposed to the cable modem. Of course I have a pentium 3 at 500Mhz and 768Mb of memory. I'm sure this helps to make it a more enjoyable experience.
reducing the number of phone lines probably wont make much of a difference. eliminating them will. besides, sometimes my isp flakes out and then I dial in so, personally, I would like at least one or two phone lines kept in service.
This response has been erased.
I have an $8/month ISP that I use to download large files once in a while, or if I really need to look at an image, or when grex is dead for five days and I have to send email. I may have used it for 2 hours this month out of the 30 I paid for. I have Opera 6 set up with Linux in case I really need to access my bank's website (they crippled it with javascript). Jim asks how many lines we are still paying for and how many are being used more than 5% of the time.
This response has been erased.
Yeah, we could put in an X11 interface that would let us remotely power cycle the terminal server. We even have all the parts. However, this hasn't been a frequent problem, so I'm too lazy to figure out how to do it. No Grex staffer or board member that I know of wants to eliminate the modems. If one show up who does, he will be pummeled into submission by the rest of the board and staff. Eliminating the modems is not an option that is being considered. Nearly everyone wants to reduce the count a bit more, but that has to wait until our centrex contract expires later this year.
I'm not sure why telnet is so slow for Sindi. It used to be for me, back before Grex got a faster 'net connection, but in the last year or two it's been pretty fast for me.
My opinions about the modems are just that, my opinions. I don't expect anyone to jump at the idea and do anything about it. It's understandable that Grex has other priorities and other goals that are not inline with mine. My post was gut reaction to Russ's attack and his not so valid argument.
This response has been erased.
We already have the hardware. Didn't I say that? In prehistoric days we used to use it to power-cycle something that kept dieing. Might have been Grex. I don't remember. But the terminal server is really pretty reliable. It was only futzed up this time because things were booted in the wrong order. A technological fix would be overkill.
This response has been erased.
This response has been erased.
I didn't ask about their dial-in options because Mary had already said that they offer the service. A separate terminal server made sense when the hardware was different. Right now, this minute, I can't decide if it still makes sense; I _like_ the idea of having the modems on a Portmaster. (I remember when a PDP-11 clone had trouble with more than six or so PPP connections at 9600bps; we've come a _long_ way since then, but I don't have a feel for the load a PPP connection would put on the main machine. Nor do I know that such a connection would make sense for grex's dial-in modems when directly connected to the grex machine.)
This response has been erased.
*I*'m suggesting we should provide PPP.
This discussion really belongs somewhere else besides the system problems item, I think. We have occasionally, over the years, discussed allowing PPP connections to Grex, because fewer and fewer computer users even know there *is* another kind of modem connection you can make. People connectiong to Grex via PPP wouldn't be allowed to http out onto the internet, but they could access Grex's web pages that way. (We don't want to compete with real ISPs.)
I'll follow-up on the modem question.
I'd be careful about using any modem that could be described as "cheap" for dial-in service. I haven't had good luck with that combination. For example, while US Robotics Sportster modems work great for dialing out, I had all kinds of reliability problems with a couple of them at work until I replaced them with Couriers. The Sportsters simply would never complete the initial negotiation if certain other brands of modem tried to dial into them.
This response has been erased.
/c is full again. from df: /dev/sd4a 1944365 1749942 0 100% /c
I dialed 7615041 and my login timed out after 120 seconds without even giving me a login prompt. Dialed 7613000 and got instant login prompt.
WARNING TO ADMINS: The failure of one port as Sindi details in #109 is exactly what presaged last week's total failure of the terminal server.
Right now, as I am dialed in, the good computer is burning a CD. The computer to grex with is the 25Mhz, 4meg of memory machine. BTW, when I telnet from M-net my pager goes screwy and pages more lines that I want. Any clue?
resp:111 - probably your terminal client isn't communicating properly what the size of your window is. Depending on what terminal you're using, you can find out how big it is --- in PuTTY, drag one of the corners a little bit, and look to the upper left corner --- and then use that number in 'stty rows <X>'. Or use the 'change' program, which basically does the same thing for you.
The following is a failed delivery message from an MSN account. Why are the at signs turning into AEA? A friend cannot get mail from her friend with MSN, he has had to retain his yahoo account to write her. > > ----- Original Message ----- > From: postmaster+AEA-mail.hotmail.com > To: eepitt2+AEA-msn.com > Sent: Saturday, July 26, 2003 11:12 AM > Subject: Delivery Status Notification (Failure) > > > This is an automatically generated Delivery Status Notification. > > Delivery to the following recipients failed. > > dpfitzen+AEA-cyberspace.org > > > > > ATTACHMENT part 2 application/octet-stream name=ATT00037.dat > ATTACHMENT part 3 message/rfc822 name=Test Message.email > From: "EDWARD PITTENGER" <eepitt2@msn.com> > To: "Dorothy Pfitzenmier" <dpfitzen@cyberspace.org> > Subject: Test Message > Date: Sat, 26 Jul 2003 11:13:57 -0500 > > This is a test message to Dorothy from my MSN E-MAIL. > > Bud
Here is more info from the person who cannot use MSN email to write grex: What is the rfc822; ? The Diagnostic-Code looks like Marcus's humor. Subject: Attachment from Bounce Back I did get this verbiage attatched to the bounce back messge, but it does not make sense to me. I hope it does to Sindi. Every thing that follows is the attachment (except for the couple of lines that YAHOO always attaches. Reporting-MTA: dns;hotmail.com Received-From-MTA: dns;mail.hotmail.com Arrival-Date: Sat, 26 Jul 2003 09:12:26 -0700 Final-Recipient: rfc822;dpfitzen@cyberspace.org Action: failed Status: 5.0.0 Diagnostic-Code: smtp;552 And he that sat upon the throne said, Behold, I make all things new. pleased to meet you 250-EXPN 250-SIZE 100000 250 HELP NING 250-8BITMIME 250-SIZE 250-DSN 250-ETRN 250-AUTH GSSAPI 250-DELIVERBY 250 HELP 250 HELP HELP 0-AUTH GSSAPI NTLM LOGIN 250-AUTH=LOGIN 250-X-LINK2STATE 250-XEXCH50 250 OK signed. which 220 have no reverse-DNS (PTR record) assigned. tions from IP addresses which 220 have no reverse-DNS (PTR record) assigned. is expressly PROHIBITED. Rinet-Turbo machines without prior WRITTEN permission 220 is expressly PROHIBITED.
RFC 822 is the document that standardised the text of Internet (at the time, Arpanet) messages.
Srw says grex is bouncing mail from this MSN account as spam and that there have been similar problems with other mail from MSN.
Received these messages when i logged in tonight: mesg: Unable to find your tty (ttyp9) in utmp file mesg: Unable to find your tty (ttyp9) in utmp file
Shortly after the previous message, about two minutes after joining party, i got the following: Party Process Killed. zsh: 16042 terminated ( mesg -h n; /usr/local/bin/party; mesg -h y ) When I tried to rejoin party, i got a message from PuTTY about my connection having been killed.
You're making ssh connections, right? At some point today Grex was out of ttys for what reason I know not. Under such circumstances ssh will connect you without a tty. This really stinks, as lots of stuff doesn't work right, like backspace keys. This may have something to do with #117. If you were connected without a tty, then robocop would have eventually killed your processes, as it dislikes all user processes not associated with a tty....yup, I just checked the log...robocop killed you.
ah. okay.
I also dial into grex about half the time (I find dial up is faster than my internet connection) but my mother's computer can only use dial up
This may not be a system problem, but for some odd reason, Backtalk has the font size set so big that it's hard to read responses.
2125 hours, dialing -3596: "Welcome to Grex! It may take a few seconds to connect." [1 minute 40 seconds later] NO CARRIER Same story on -3000 a few minutes later. I'm assuming Grex was down.
Yup. 'Twas being switched back to the UPS.
re 122 Wow, when will you know if it's really a system problem? Perhaps you set your fonts really big on your browser.
Backtalk looks fine to me, jaklumen, maybe you should check your browser, like naftee suggested. Hi Naftee!!
That would be like >View -> Text Size -> Medium.
resp:117--119 - had this same problem again last night.
Thanks-- oddly enough, I wasn't seeing the problem well enough on other sites.
I can at least say that nothing in Backtalk has changed. It mostly avoids fiddling with your fonts, except that it tries to force your browser to use a monospaced font in input boxes. This has to be a browser issue of some sort.
I'm fine, now... other sites had a few fonts looking a bit biggish. I wonder how I accidentally set it to "Largest."
I think there's a hotkey combination (When s certain combination of your keyboard keys will behave like a mouse click on a selection) that does that. I've done that a couple of times by mistake. I've also done something that makes my toolbar in all my MS applications appear really big, but I don't know what I did, and don't have the time or inclination to figure out how to fix it. GOtta love those "hotkeys"
Took about 3 minutes to get a login prompt, and just recieved a message out of the blue that "grex.cyberspace.org does not seem to exist", or words to that effect.
I can't send mail either.
8:45pm up 2 days, 22:58, 2 users, load average: 0.21, 0.18, 0.01 Login Name TTY Idle Login Time Location Work Phone penyair ravi shadanah p0 3:58 Aug 1 02:41 rcurl Rane Curl *u1 Aug 1 20:45
Another extremely long delay for a login prompt, and mail from a very reliable source is not getting through. The DSL line is up, but only one user is coming in from the Internet. Maybe DNS is down; this is definitely bad news.
Outbound telnet and ftp don't seem to do anything either. (Maybe they also have the 3-minute delay; I'm not that patient.)
It looks like there was a network problem. I don't have any more information than: I couldn't get to grex and traceroute was failing at Voyager, our ISP.
Turns out the DSL modem needed a power cycle. The internet user Russ saw, possibly the terminal server?
Maybe we should get the DSL modem onto the UPS.
We took a power hit, long enough to power down the computers. We aren't too far away from the pumpkin. The problems started at that time. Grex went down, then up, then down. Looked sort of like the power went down with the UPS taking over, then the power coming back up.
Well, 'grex' didn't go down; it's been up three days, four hours and thirty-nine minutes so far. But I can believe _something_ went down due to power.
Well, just glad Grex is back. Was worried there for a bit.
Doctor Clu
/|\ TT 030
Yeah arbornet just died. Hi mynxcat!!
Hi naftee!
I'm sure this is indicative of a problem (two users, one pty?): niqu miki dude p0 6:07 Aug 4 01:30 penyair ravi shadanah p0 6:07 Aug 1 02:41
??? 3 days apart?
Haha
Russ, that's a harmless bug which has been DISCUSSED TO DEATH SEVERAL FUCKING TIMES ALREADY. :)
scott, please. drop the caps and FUCKING CONTROL YOURSELF"
scott can raise is voice WHENEVER HE FUCKING FEELS LIKE IT.
charcat grabs the squirty water bottle thingie and squirts all people using caps! =^o.-^=
I don't seem to be receiving mail. I know of at least two messages that have been sent to me in the last day, but mailx still returns "No mail for dcat", and mutt and pine complain about a non-existant mailspool file (which would be normal if I didn't have mail in it, but I do). I've gotten the no mailspool file errors from PINE & Mutt before, but they went away when I got new mail; this doesn't seem to be happening this time. And no, I don't have a .forward file.
Hmm... } !ls -l /var/spool/mail/d/c/dcat } -rw------- 1 dcat cohorts 9985 Aug 12 12:30 } /var/spool/mail/d/c/dcat Looks like you have something now.
Re #153: Are you using ssh or telnet? SSH doesn't always set the MAIL variable right, leaving mailx and Pine unable to find your spool file. If you have mail when you telnet in, but not when you ssh, add a line to your .profile to set MAIL to whatever it's set to when you telnet in. That's what I had to do. This problem really puzzled me for a while, too.
Thank you Gull, that does seem to have done it. Now I just have to figure out what file to put that in to have it read in time for the 'You have new mail' message at logon. .zshenv (the normal place for it such things in zsh, IIRC) doesn't seem to be early enough.
For CSH, the order of execution is .cshrc and then .login
I SSH in and have had no problem with the MAIL variable getting set. It might be the shell failling to set it. Still, good advice.
Could be. I use bash as my shell, and I had the problem.
The net connection seems to be still down.
(actually, it appears to be a DNS problem; I'm currently on via telnet, having used the IP address.)
It does indeed appear to be a DNS problem; I can make connections from grex to the outside world. (I'm in the pumpkin right now.) FWIW, named is running on grex.
Oh. I forgot to mention: one of grex's secondary name servers appears to be off the net itself right now. The other one seems to be having a different problem, but I don't know what that problem is: it has grex's NS records, but not the address, mail exchange or other records. I suspect that both are suffering effects of the recent blackout.
/tmp became full a few minutes here before 3pm. That means e-mail replies can't be sent...
/a is full.
Unlike your BRAIN.
Got dropped twice earlier this evening, calling -3596. Also, some asshole has filled /a to the brim.
It's still "almost" full. There's about 2.5 megs free.
Use them and quit complaining.
You would say that. You're lucky enough to have a home directory on /d.
Dear Grex:
What's with the frellin' hangin' up of the phone when I dial in?
I've seen this too, sometimes, recently. Had to give up & dial in later.
This response has been erased.
re 170 I never noticed, but thanks for pointing it out
polytarp's password has been changed.
sometimes green text showed up on my screen yesterday. Then it went away, this happened about 4 times, two different sessions.
re #176: that's likely because polytarp's conferencing name has been changed to something which contains terminal-control escape sequences which select a different font color..
U.p.
/a is full again. I created another loginid in the hopes I'd get space on one of the home partitions that doesn't fill up quite so often, but it stuck that one on /a too. :P
I got polytarp's green too. I figured it was his as it showed up simultaneously with a response from him. What would have been the simplest online way to neutralize it? I logged off and back in.
Hmmm . . . I don't know about 'simplest', but you could set your terminal software to not execute escape sequences. Unfortunately, I don't know how to do that; my terminal was not affected by that little . . . joke.
I'm not sure if this is a problem or by design, but when I try to ftp to Grex to download files, my ssh connection gets dropped almost instantly.
Why not use scp in the first place?
I have a problem completely unrelated to the disk-eating assholes: I'm having a difficult (read: impossible) time trying to transfer files to and from Grex. First, FTP seems to be blocked. Hard. Here is the result of two attempts to log in to Grex; in both cases ftp simply froze until I hit ^\, no user prompt, no password prompt: [russ@localhost russ]$ ftp cyberspace.org Connected to cyberspace.org (216.93.104.34). Quit [russ@localhost russ]$ ftp cyberspace.org Connected to cyberspace.org (216.93.104.34). Quit A trial of ftp from ftp.sri.com immediately after this worked fine. This proves that there is no problem between my keyboard and the greater Internet; it is a Grex problem. (M-Net doesn't work either, but it lets me log in and only fails when I try to transfer data.) Second, I've been trying to tunnel ftp through ssh in order to get around these blocks. I forwarded port 8021 to Grex port 21, and got a reward: [russ@localhost russ]$ ftp ftp> open localhost 8021 Connected to localhost (127.0.0.1). 220 grex.cyberspace.org FTP server (Version wu-2.6.1-GREX(10) Mon Dec 3 00:09:59 EST 2001) ready. Name (localhost:russ): russ 331 Password required for russ. Password: 230->>>NO PSYBNC<<< >>>NO EGGDROP<<< NO NO NO!!!!! Won't run here! 230- [much deleted] 230 User russ logged in. Remote system type is UNIX. Using binary mode to transfer files. ftp> binary 200 Type set to I. ftp> get agora.Z.26 local: agora.Z.26 remote: agora.Z.26 227 Entering Passive Mode (216,93,104,34,247,62) 425 Possible PASV port theft, cannot open data connection. ftp> So there you go. I can run a terminal session over TCP/IP, but I can't pull down copies of my mail or anything else of use. What do I have to do to get the functionality of a lousy dial-in user?
Dial in? ;-)
This response has been erased.
Interesting. I'd tried scp but got error messages and didn't see how to relate them to corrective action, so I gave up. Tried just now and it actually works. It's clunky but I think I can manage. Thanks!
The /a partition was full, preventing users whose home directories are located on it from responding in bbs, saving their mail, and doing anything else requiring writing data to the disk. I identified the major disk hogs and did some cleaning up, giving us a bit of breathing room.
[Loud cheer for the Grexasaurus from the /a users]
I've had no trouble ftp-ing with the client Fetch over TCP/IP.
re 177 polytarphs account got spalttered?
/a was full a few minutes ago. It's now got a little space, hopefully enough to let me enter this response without crashing something.
Like your brain?
A friend with an email account at mymailbox.com reports that every mail he gets from grex (but not my ISP), including my account and jdeigert, arrives in 4-7 copies. What might be causing this? Other people don't tell me they get multiple copies of mail from me.
myrealbox.com not mymailbox.com
He should look at the full headers of the copies, to compare the lines that begin "Received:" The ones at the bottom should be the same in every copy, but then there will be some that have different time-stamps. Those lines will show where the transfer is failing. Most likely, a machine passes it on, but then does not get the acknowledgement of receipt, so it re-queues the message, to try again later. The machine that received, but did not acknowledge, the message delivers it. Result: the message is duplicated when delivery is attempted later.
No wonder sendmail uses so much CPU all the time.
Very oddly, when ssh'ing to Grex I only *sometimes* get the message that Grex is lying about the size of its private key. Is there any mechanism I could use to confirm the key and rule out funny business?
I'm not sure what it is, but I had some problems with Backtalk... I had to put my password in several times at a few points to get in and through the bbs.
Yes, I'm getting the same problem. I just gave up on Backtalk after encountering it several times. If I exited from all of my IE windows, then started one up again and logged into Grex, I could read about 3 items before getting it again.
I've noticed that happens a lot when Grex's CPU load is very high. I suspect something is timing out.
This response has been erased.
re 184: i would suggest using kermit or x/y/z-modem across your tcp connection, but that appears reserved for dial-up users (foolishly so imo).
A response when Jamie begs the Grexers for some loving: Would a kind Grexer with the rootkey please reset the password for jp2? To authenticate me, please call the number shown in !f jp2. Thank you, <hug>
This response has been erased.
Try reading (or running, if you're telnetted in) /usr/local/bin/sz, & you'll see.
Send mail to staff@cyberspace.org, Jamie.
Aww, Jamie hugs are the best. Jamie, try emailing staff with the request.
Traditionally none of those protocols have worked well over telnet because telnet is not transparent, and often not 8-bit clean. ssh is, if you disable the escape character with "-e none".
What does it mean when only certain people who try to email me get "permanent fatal error" messages?
Are they getting error messages with bible quotes in them?
That you're a femme fatale? :)
When I have seen those bounces, it was usually due to a temporary connection problem, "temporary" being variable...
I emailed the staff on like, Aug 29.
(It looks like /var/spool/mail is a bit full.)
Filesystem kbytes used avail capacity Mounted on /dev/sd3h 1944365 1754424 0 100% /var/spool/mail
Filesystem kbytes used avail capacity Mounted on /dev/sd3h 1944365 1756006 0 100% /var/spool/mail /dev/sd7g 1971009 1773911 0 100% /d 2 partitions that are widely used are full :( 2 that i use :(
I thought you left?
I THOUGHT YOU WERE A WOMAN
I've never seen /var/spool/mail full before, so I had to improvise. I found some files I could safely delete or move to temporarily make some space. Valerie started a reap, which should end up deleteing half the accounts on Grex (it's been a long time since we ran a reap) and making plenty of space everywhere. The shortage of space on /d was pretty much just one user.
resp:218 That's another problem i have.
re 220 You should run reaps more often. Or buy more hard drives.
Grex should consider upgrading its ancient version of Pine. Versions earlier than 4.57 have a remotely exploitable security hole.
and use nano instead of pico.
Haven't struggled with it myself, but my impression is that updating Pine is non-trivial on our ancient OS. Installing "nano" shouldn't be a big deal. I'll look at it when I get time, unless some other staffer beats me to it.
Use elm. It's the best.
Why not use !mail ?
Yeah, why not?
I have been a supporter of Grex in the past by having several small non-profit organizations with which I have been associated join Grex and use it at least as their website and board mail reflector. The latter, however, has become untenable because of spam. There is nearly ten times more spam being distributed to the boards than board correspondence. Is there any hope of soon having access to a filter here for spam? I will probably move an organization off Grex (and thereby cancel membership) unless there is some recourse against this avalanche of junk e-mail.
It may not get better elsewhere. My Grex account is actually on the low end as far as the amount of spam I get. My ameritech.net account gets over 50 spams a day, and has since before I started using it! My work account gets about 100 a day, about 90% of which is caught by a statistical filter. I agree that Grex needs better spam filtering, but I want the ability to turn it off if I choose to. I've had too many bad experiences with mail disappearing because of spam filters -- no bounce message or anything, just disappearing into the ether with no warning.
(NB: This is the System Problems item from Summer, 2003, not Fall, 2004, so it doesn't have quite the audience the item in the current agora would get.)
(I don't think Fall 2004 will have much audience - yet - either.......but thanks for the poke in the ribs 8^})
Could Grex use the Spamhaus Block List (SBL) to block spam? See http://www.spamhaus.org/sbl/howtouse.html
We will have to put a kludge in it as cyberspace.org sometimes ends up on various spamblock lists. Might be worth considering though.
I'm telnetting in from Madeira Beach FL. I cannot connect to Grex directly as I get a "not responding" response, but I have telnetted into CAEN, and then telnetted over from there. Why won't Grex respond directly?
Grex is responding very slowly. I'm not surprised you are having trouble.
In fact the problems are so bad that this report got stored in *old* agora! ;-)
You have several choices: