From bernard at blackham.com.au Mon Mar 6 20:05:15 2006 From: bernard at blackham.com.au (Bernard Blackham) Date: Mon, 6 Mar 2006 20:05:15 +0800 Subject: [tech] axis building Message-ID: <20060306120515.GD18471@blackham.com.au> Notes for archiving... this is what you need to build dropbear 0.47. export PATH=/usr/local/cris:$PATH export CFLAGS="-melinux -O3 -s -symbolic -D__uClinux__" ./configure --prefix=/mnt/flash/dropbear --disable-zlib --host=cris make MULTI="1" STATIC="1" PROGRAMS="dropbear dbclient scp" cris-gcc -melinux -s -symbolic -static -o dropbearmulti dbmulti.o atomicio.o bignum.o buffer.o circbuffer.o cli-algo.o cli-auth.o cli-authinteract.o cli-authpasswd.o cli-authpubkey.o cli-channel.o cli-chansession.o cli-kex.o cli-main.o cli-runopts.o cli-service.o cli-session.o cli-tcpfwd.o common-algo.o common-channel.o common-chansession.o common-kex.o common-runopts.o common-session.o compat.o dbutil.o dss.o fake-rfc2553.o listener.o loginrec.o packet.o process-packet.o progressmeter.o queue.o random.o rsa.o scp.o scpmisc.o signkey.o sshpty.o svr-agentfwd.o svr-algo.o svr-auth.o svr-authpam.o svr-authpasswd.o svr-authpubkey.o svr-chansession.o svr-kex.o svr-main.o svr-runopts.o svr-service.o svr-session.o svr-tcpfwd.o svr-x11fwd.o tcp-accept.o termcodes.o libtomcrypt/libtomcrypt.a libtommath/libtommath.a -L /usr/local/cris/lib/gcc-lib/cris/2.96/elinux/ uccrt/*.o There's custom stuff in uccrt/ which should also be archived somewhere. Doesn't work quite yet - needs some hacking to get around a buggy libc. Bernard. -- Bernard Blackham From grahame at angrygoats.net Sun Mar 19 01:55:54 2006 From: grahame at angrygoats.net (Grahame Bowland) Date: Sun, 19 Mar 2006 01:55:54 +0800 Subject: [tech] Phonehome Message-ID: Hi guys I've just set up some software I developed in spare time for ITS use. It's a pretty simple system I've named "Phonehome". On the server side (mooneye) a page to show the results, and a page to accept updates via SOAP. http://phonehome.ucc.gu.uwa.edu.au/ Plugins are trivial. They should be written (and tested) on the server, see /usr/local/phonehome/client/plugins on mooneye. When you're happy a plugin should be pushed out, cd /usr/local/phonehome && ./updateall.zsh - you'll need to forward an SSH agent through or whatever though. The client is run out of cron.daily overnight. Current plugins; - run apt-get update && apt-get dist-upgrade -d -u, which downloads any new packages but doesn't install them. Then uses apt-listchanges to determine the necessity of the upgrades. - bunches of informational plugins - there's a plugin to check afbackup and make sure dumps are happening; someone should write something to check up on amanda. ... Etc, etc. I've found this really helpful at ITS, as I just check it routinely to see if there are any important updates my machines are missing. Same goes for UCC; it'll be useful as long as it gets looked at. Anyway, if there are any questions just mail me. If a host shows up as "possibly down" it means the last time it gave any results was > 48 hours ago. Oh, and someone more confident they won't break stuff should go through and do the security updates it has found. Access to the site is currently limited by IP, you pretty much need to be on 130.95.13.0/24 to get in. Someone could set up user auth if they felt like it, just make sure not to break access to the service.py script :-) There is a password in /usr/local/phonehome/etc/client.pass which should be different on most machines (with exception of pitch/velvet.) That gets cross-referenced against /usr/local/phonehome/etc/passphrases on mooneye. Anyway, hopefully it'll be useful. Cheers Grahame From trs80 at ucc.gu.uwa.edu.au Mon Mar 20 23:12:57 2006 From: trs80 at ucc.gu.uwa.edu.au (James Andrewartha) Date: Mon, 20 Mar 2006 23:12:57 +0800 (WST) Subject: [tech] bronze (the alloy switch) Message-ID: Grahame helped me poke bronze into exporting its management interface over a trunked vlan. The trick is to set the PVID to 2 on any trunked port (the management interface is on vlan 2). Anyway, mussel, martello, madako, olive and the uplink are plugged into it: madako: e1000: eth0: e1000_watchdog_task: NIC Link is Up 1000 Mbps Full Duplex martello: tg3: eth1: Link is up at 1000 Mbps, full duplex. In unrelated news, maroon's soundcard plays for about 2 seconds and then sets everything touching it to the D state. -- # TRS-80 trs80(a)ucc.gu.uwa.edu.au #/ "Otherwise Bub here will do \ # UCC Wheel Member http://trs80.ucc.asn.au/ #| what squirrels do best | [ "There's nobody getting rich writing ]| -- Collect and hide your | [ software that I know of" -- Bill Gates, 1980 ]\ nuts." -- Acid Reflux #231 / From zanchey at ucc.gu.uwa.edu.au Tue Mar 21 16:10:15 2006 From: zanchey at ucc.gu.uwa.edu.au (David Adam) Date: Tue, 21 Mar 2006 16:10:15 +0800 (WST) Subject: [tech] bronze (the alloy switch) In-Reply-To: References: Message-ID: James, On Mon, 20 Mar 2006, James Andrewartha wrote: > Grahame helped me poke bronze into exporting its management interface over > a trunked vlan. The trick is to set the PVID to 2 on any trunked port (the > management interface is on vlan 2). Anyway, mussel, martello, madako, > olive and the uplink are plugged into it: Can you update the appropriate documentation? > In unrelated news, maroon's soundcard plays for about 2 seconds and then > sets everything touching it to the D state. I queued something this morning to test it, it entered the D state and then about three hours later continued to play. It seems to be fine now - did anyone change anything? David Adam UCC Wheel Member & 32nd President zanchey@ From grahame at angrygoats.net Tue Mar 21 17:43:41 2006 From: grahame at angrygoats.net (Grahame Bowland) Date: Tue, 21 Mar 2006 17:43:41 +0800 Subject: [tech] Mussel reconfigured Message-ID: <20060321094341.GA27617@angrygoats.net> Hey guys Some quick changes made with the help of Davyd. eth0 ip moved to eth0.2, along with all the eth0:1, eth0:2.. interfaces. changed port 3 on bronze to VLAN trunk, sending over vlans 2 and 7. Printing now works again. I think that the text file network description is antiquated. Much better off using the switches themselves as the doco and using snmp. Guess can't realy do that as much with the alloy, but it's nice being able to do sh int Fa0/2 and find out that X is on it from the description.. Cheers Grahame From grahame at angrygoats.net Wed Mar 22 11:51:22 2006 From: grahame at angrygoats.net (Grahame Bowland) Date: Wed, 22 Mar 2006 11:51:22 +0800 Subject: [tech] FW: WA Cisco Stock In-Reply-To: <006301c64d4c$15414490$0201010a@mycomp> Message-ID: Hi guys Just got this quote from Computer Resale Brokers WA. I'm sure ITS has bought from them in the past.. Anyway, probably about the cheapest way to plug manbo into the Gigabit switch. We'll also need a patch lead, I'm sure Mark can buy one of those cheaply through Page Data or something.. So, we're looking at something around $500 to plug it in at Gig. Sounds like a fundraising project to me.. I'm in for $50, anyone else? :-) (Come on, gig fibre links in the machine room are Cool damnit.) Cheers Grahame ------ Forwarded Message From: "Computer Resale Brokers (WA)" Date: Wed, 22 Mar 2006 09:00:57 +0800 To: Grahame Bowland Subject: Re: WA Cisco Stock Hi Grahame, Thanks for your price request. We can supply and we have in stock the following modules: GLC-SX-MM................................................................... ...........................$300ea (GE SFP, LC connector SX transceiver) WS-G5484.................................................................... ................................$190ea 1000base SX GBIC (multimode) (prices for single order) Note: All prices are ex-GST and delivery. All GBIC'S and SFP's New Compatible. All New Agilent and Prolab GBiC's 3 years warranty Regards, John Bieganski Computer Resale Brokers Ph/Fax (08) 9344 5605 Mob 0412 284 456 Email: john.bieganski at computer-resale-brokers.com http://www.computer-resale-brokers.com ----- Original Message ----- From: "Grahame Bowland" To: "Computer Resale Brokers (WA)" Sent: Tuesday, March 21, 2006 6:29 PM Subject: Re: WA Cisco Stock > Hello > > I'm interested in indicitive pricing on a single SX GBIC and also on a > single SX SFP module. > > Thanks, > Grahame > > > > > > -- > No virus found in this incoming message. > Checked by AVG Free Edition. > Version: 7.1.385 / Virus Database: 268.2.6/286 - Release Date: 3/20/2006 > > ------ End of Forwarded Message From trs80 at ucc.gu.uwa.edu.au Wed Mar 22 12:07:24 2006 From: trs80 at ucc.gu.uwa.edu.au (James Andrewartha) Date: Wed, 22 Mar 2006 12:07:24 +0800 (WST) Subject: [tech] FW: WA Cisco Stock In-Reply-To: References: Message-ID: On Wed, 22 Mar 2006, Grahame Bowland wrote: > Just got this quote from Computer Resale Brokers WA. I'm sure ITS has bought > from them in the past.. Anyway, probably about the cheapest way to plug > manbo into the Gigabit switch. We'll also need a patch lead, I'm sure Mark > can buy one of those cheaply through Page Data or something.. > > So, we're looking at something around $500 to plug it in at Gig. Sounds like > a fundraising project to me.. I'm in for $50, anyone else? :-) > > (Come on, gig fibre links in the machine room are Cool damnit.) Um, manbo already has a fibre gigE interface, so all we really need is a SX SFP for bronze, which come in at $120 or so https://www.itfast.com.au/shopnow.asp?ID=MGBIC-MLC -- # TRS-80 trs80(a)ucc.gu.uwa.edu.au #/ "Otherwise Bub here will do \ # UCC Wheel Member http://trs80.ucc.asn.au/ #| what squirrels do best | [ "There's nobody getting rich writing ]| -- Collect and hide your | [ software that I know of" -- Bill Gates, 1980 ]\ nuts." -- Acid Reflux #231 / From davyd at madeley.id.au Wed Mar 22 12:47:12 2006 From: davyd at madeley.id.au (Davyd Madeley) Date: Wed, 22 Mar 2006 12:47:12 +0800 Subject: [tech] FW: WA Cisco Stock In-Reply-To: References: Message-ID: <1143002832.5117.17.camel@frobisher.madeley.id.au> On Wed, 2006-03-22 at 12:07 +0800, James Andrewartha wrote: > On Wed, 22 Mar 2006, Grahame Bowland wrote: > > > Just got this quote from Computer Resale Brokers WA. I'm sure ITS has bought > > from them in the past.. Anyway, probably about the cheapest way to plug > > manbo into the Gigabit switch. We'll also need a patch lead, I'm sure Mark > > can buy one of those cheaply through Page Data or something.. > > > > So, we're looking at something around $500 to plug it in at Gig. Sounds like > > a fundraising project to me.. I'm in for $50, anyone else? :-) > > > > (Come on, gig fibre links in the machine room are Cool damnit.) > > Um, manbo already has a fibre gigE interface, so all we really need is a > SX SFP for bronze, which come in at $120 or so > https://www.itfast.com.au/shopnow.asp?ID=MGBIC-MLC Which we're going to trade for $3000 in kit. Cool! --d -- Davyd Madeley http://www.davyd.id.au/ 08B0 341A 0B9B 08BB 2118 C060 2EDD BB4F 5191 6CDA From zanchey at ucc.gu.uwa.edu.au Fri Mar 24 13:23:45 2006 From: zanchey at ucc.gu.uwa.edu.au (David Adam) Date: Fri, 24 Mar 2006 13:23:45 +0800 (WST) Subject: [tech] Power outage Message-ID: The power to the Guild went out about 10:15am this morning. As usual, not everything came back up smoothly (morwong took a couple of goes, mussel's network did odd things, madako required ifup/downing, a whole stack of things that we shouldn't have to do). I know there were a couple of wheel members doing things as well as me, but unfortunately I have no idea what they were doing or what they did. Most things should be back now - e-mail or call the clubroom if you have problems. David Adam UCC Wheel Member From matt at ucc.asn.au Fri Mar 24 17:38:30 2006 From: matt at ucc.asn.au (Matt Johnston) Date: Fri, 24 Mar 2006 17:38:30 +0800 Subject: [tech] mermaid apache2 upgrade Message-ID: <20060324093830.GA16910@ucc.gu.uwa.edu.au> Hi all. I've upgraded the webserver on mermaid to apache2. Everything I've looked at seems working, though if you come across anything broken let me know. Cheers, Matt From nick at ucc.gu.uwa.edu.au Fri Mar 24 23:18:31 2006 From: nick at ucc.gu.uwa.edu.au (Nick Bannon) Date: Fri, 24 Mar 2006 23:18:31 +0800 Subject: [tech] Coke's down Message-ID: <20060324151829.GA20858@ucc.gu.uwa.edu.au> Anyone know what it takes to bring it up? Standard tricks involve "rmmod 8250; modprobe 8250", running minicom to set up the serial port (the coke machine is working fine interactively). In response to the serial port writes, multiple blank lines are being read back in... Nick. mermaid:/home/wheel/nick# strace -s 48 dispense finger coke [...] open("/dev/ttyS0", O_RDWR) = 3 flock(3, LOCK_EX|LOCK_NB) = 0 ioctl(3, SNDCTL_TMR_TIMEBASE or TCGETS, {B9600 opost isig icanon -echo ...}) = 0 [...] ...}) = 0 write(3, "d7\n", 3) = 3 read(3, "d7\n", 100) = 3 read(3, "\n", 100) = 1 read(3, "ok\n", 100) = 3 open("/var/coke/coke.conf", O_RDONLY) = 5 fstat64(5, {st_mode=S_IFREG|0644, st_size=506, ...}) = 0 old_mmap(NULL, 4096, PROT_READ|PROT_WRITE, MAP_PRIVATE|MAP_ANONYMOUS, -1, 0) = 0 read(5, "slots 7 16\nslot 0 86 null coke\nslot 1 73 lemonad"..., 4096) = 506 write(3, "n0 null coke\n", 20) = 20 read(3, "\n", 100) = 1 read(3, "\n", 100) = 1 read(3, "\n", 100) = 1 write(3, "n1 lemonade\n", 20) = 20 read(3, ":\n", 100) = 2 read(3, "\n", 100) = 1 read(3, ":n0 null coke\n", 100) = 21 write(3, "n2 screaming soda\n", 20) = 20 read(3, "\n", 100) = 1 read(3, "ok\n", 100) = 3 read(3, "\n", 100) = 1 [...] munmap(0xb7fe8000, 4096) = 0 write(3, "s\n", 2) = 2 read(3, ":n2mg soda\n", 100) = 11 read(3, "\n", 100) = 1 read(3, "ok\n", 100) = 3 read(3, "\n", 100) = 1 read(3, "\n", 100) = 1 read(3, "\n", 100) = 1 read(3, ":\n", 100) = 2 read(3, "\n", 100) = 1 read(3, ":n3lon foo\n", 100) = 11 flock(3, LOCK_UN) = 0 close(3) = 0 fstat64(1, {st_mode=S_IFCHR|0620, st_rdev=makedev(136, 1), ...}) = 0 old_mmap(NULL, 4096, PROT_READ|PROT_WRITE, MAP_PRIVATE|MAP_ANONYMOUS, -1, 0) = 0 write(1, "The UCC Coke machine.\n", 22) = 22 write(1, "\n", 1) = 1 write(1, "Slot 0 has sold 922 drinks of type \204\372\377\277.\n", 41) = 41 write(1, "Slot 1 has sold 106 drinks of type \204\372\377\277.\n", 41) = 41 [...] Nick. -- Nick Bannon | "I made this letter longer than usual because nick-sig at rcpt.to | I lack the time to make it shorter." - Pascal From nick at ucc.gu.uwa.edu.au Sat Mar 25 00:15:24 2006 From: nick at ucc.gu.uwa.edu.au (Nick Bannon) Date: Sat, 25 Mar 2006 00:15:24 +0800 Subject: [tech] Coke's up In-Reply-To: <20060324151829.GA20858@ucc.gu.uwa.edu.au> References: <20060324151829.GA20858@ucc.gu.uwa.edu.au> Message-ID: <20060324161524.GB20858@ucc.gu.uwa.edu.au> On Fri, Mar 24, 2006 at 11:18:31PM +0800, Nick Bannon wrote: > Anyone know what it takes to bring it up? > > Standard tricks involve "rmmod 8250; modprobe 8250", running minicom to > set up the serial port (the coke machine is working fine interactively). [...] ...and [DAG]'s magic touch brought it back! mermaid:~# killall dispense && minicom ... hit enter a couple of times ... && dispense standalone coke mmm, i sometimes have to type n Nick. -- Nick Bannon | "I made this letter longer than usual because nick-sig at rcpt.to | I lack the time to make it shorter." - Pascal From grahame at angrygoats.net Sat Mar 25 20:40:14 2006 From: grahame at angrygoats.net (Grahame Bowland) Date: Sat, 25 Mar 2006 20:40:14 +0800 Subject: [tech] Manbo; maroon Message-ID: I moved manbo on to the gig switch. This didn't fix the NFS issues for maroon, eventually tracked down NFS having stupid buffer sizes set, so that manbo kept sending enormous fragmented packets over. rsize=1024,wsize=1024 fixed it (/etc/fstab) All good now. Grahame From zanchey at ucc.gu.uwa.edu.au Sat Mar 25 20:56:46 2006 From: zanchey at ucc.gu.uwa.edu.au (David Adam) Date: Sat, 25 Mar 2006 20:56:46 +0800 (WST) Subject: [tech] Manbo; maroon In-Reply-To: References: Message-ID: On Sat, 25 Mar 2006, Grahame Bowland wrote: > I moved manbo on to the gig switch. This didn't fix the NFS issues for > maroon, eventually tracked down NFS having stupid buffer sizes set, so that > manbo kept sending enormous fragmented packets over. > > rsize=1024,wsize=1024 fixed it (/etc/fstab) > > All good now. The Ubuntu song may now sound across the plains and prairies of Cameron Hall. Also on Manbo, I upgraded a stack of the CSW/Blastwave packages today, so if things suddenly start breaking let me know - I find it unlikely, though. David Adam UCC Wheel Member From elixxir at ucc.asn.au Sun Mar 26 09:28:41 2006 From: elixxir at ucc.asn.au (Paul Marinceu) Date: Sun, 26 Mar 2006 09:28:41 +0800 Subject: [tech] martello ssh dropping out on idle Message-ID: <20060326012841.GA8087@ucc.gu.uwa.edu.au> Hi, Anyone know what is the cause of this. It drops the session on about 3-5 minutes of idle time. Is this by design? (don't think so as it's pretty stupid) Had a look at sshd_config and it's just a standard config, no mods. So maybe it's the router? Any clues? PS. This has been going on for ages it's just I kept hoping it'll get fixed by someone else :P From davyd at madeley.id.au Sun Mar 26 09:30:32 2006 From: davyd at madeley.id.au (Davyd Madeley) Date: Sun, 26 Mar 2006 09:30:32 +0800 Subject: [tech] martello ssh dropping out on idle In-Reply-To: <20060326012841.GA8087@ucc.gu.uwa.edu.au> References: <20060326012841.GA8087@ucc.gu.uwa.edu.au> Message-ID: <20060326013032.GB22224@bridgewayconsulting.com.au> On Sun, Mar 26, 2006 at 09:28:41AM +0800, Paul Marinceu wrote: > Hi, > > Anyone know what is the cause of this. It drops the session on about 3-5 > minutes of idle time. Is this by design? (don't think so as it's pretty > stupid) > > Had a look at sshd_config and it's just a standard config, no mods. So maybe > it's the router? > > Any clues? > > PS. This has been going on for ages it's just I kept hoping it'll get fixed > by someone else :P I've not seen this. Have you seen this on any other hosts? --d -- Davyd Madeley http://www.davyd.id.au/ 08B0 341A 0B9B 08BB 2118 C060 2EDD BB4F 5191 6CDA From elixxir at ucc.asn.au Sun Mar 26 09:35:33 2006 From: elixxir at ucc.asn.au (Paul Marinceu) Date: Sun, 26 Mar 2006 09:35:33 +0800 Subject: [tech] martello ssh dropping out on idle In-Reply-To: <20060326013032.GB22224@bridgewayconsulting.com.au> References: <20060326012841.GA8087@ucc.gu.uwa.edu.au> <20060326013032.GB22224@bridgewayconsulting.com.au> Message-ID: <20060326013533.GA8377@ucc.gu.uwa.edu.au> On Sun, Mar 26, 2006 at 09:30:32AM +0800, Davyd Madeley wrote: > I've not seen this. Have you seen this on any other hosts? Hmm. Ok, I'll try mussel or manbo (I'm always on martello so haven't tried any others but I'm surprised no one else experinced this). This happens to me from multiple locations so it's definitely not my end. From trs80 at ucc.gu.uwa.edu.au Sun Mar 26 12:39:39 2006 From: trs80 at ucc.gu.uwa.edu.au (James Andrewartha) Date: Sun, 26 Mar 2006 12:39:39 +0800 (WST) Subject: [tech] martello ssh dropping out on idle In-Reply-To: <20060326013533.GA8377@ucc.gu.uwa.edu.au> References: <20060326012841.GA8087@ucc.gu.uwa.edu.au> <20060326013032.GB22224@bridgewayconsulting.com.au> <20060326013533.GA8377@ucc.gu.uwa.edu.au> Message-ID: On Sun, 26 Mar 2006, Paul Marinceu wrote: > On Sun, Mar 26, 2006 at 09:30:32AM +0800, Davyd Madeley wrote: >> I've not seen this. Have you seen this on any other hosts? > > Hmm. Ok, I'll try mussel or manbo (I'm always on martello so haven't tried > any others but I'm surprised no one else experinced this). > > This happens to me from multiple locations so it's definitely not my end. I'm seeing it too, but going to morwong, and I'm definitely not idle - it drops out while I'm typing stuff. -- # TRS-80 trs80(a)ucc.gu.uwa.edu.au #/ "Otherwise Bub here will do \ # UCC Wheel Member http://trs80.ucc.asn.au/ #| what squirrels do best | [ "There's nobody getting rich writing ]| -- Collect and hide your | [ software that I know of" -- Bill Gates, 1980 ]\ nuts." -- Acid Reflux #231 / From bernard at blackham.com.au Sun Mar 26 13:40:10 2006 From: bernard at blackham.com.au (Bernard Blackham) Date: Sun, 26 Mar 2006 13:40:10 +0800 Subject: [tech] martello ssh dropping out on idle In-Reply-To: <20060326012841.GA8087@ucc.gu.uwa.edu.au> References: <20060326012841.GA8087@ucc.gu.uwa.edu.au> Message-ID: <20060326054010.GC14328@blackham.com.au> On Sun, Mar 26, 2006 at 09:28:41AM +0800, Paul Marinceu wrote: > Anyone know what is the cause of this. It drops the session on about 3-5 > minutes of idle time. Is this by design? (don't think so as it's pretty > stupid) > > Had a look at sshd_config and it's just a standard config, no mods. So maybe > it's the router? Can't say I've ever encountered this. I've left ssh connections to most UCC machines open and idle all night, or all weekend, and none have died. If it's only when it's idle, then perhaps you can try setting ServerAliveInterval to 60 on your end? Though if it is as James has seen, and it can happen at any moment, it's time to pull out tcpdump and see who sends what. Bernard. -- Bernard Blackham From elixxir at ucc.gu.uwa.edu.au Mon Mar 27 09:18:51 2006 From: elixxir at ucc.gu.uwa.edu.au (elixxir@ucc.gu.uwa.edu.au) Date: Mon, 27 Mar 2006 09:18:51 +0800 Subject: [tech] martello ssh dropping out on idle In-Reply-To: <20060326054010.GC14328@blackham.com.au> References: <20060326012841.GA8087@ucc.gu.uwa.edu.au> <20060326054010.GC14328@blackham.com.au> Message-ID: <20060327091851.stzbzngjkwg88w8c@secure.ucc.asn.au> Quoting Bernard Blackham : > Can't say I've ever encountered this. I've left ssh connections to > most UCC machines open and idle all night, or all weekend, and > none have died. Yeah, for me it's happening on all UCC boxen, martello, mussel and manbo. Will try to look into it with tcpdump. ---------------------------------------------------------------- This message was sent using IMP, the Internet Messaging Program.