TheGrebs.com:IRC Logs:#linode:2007:01:23

---	Log	opened Tue Jan 23 00:00:29 2007
00:22	\|-\|	erikh [~erikh@hollensbe.org] has joined #linode
00:22	<taupehat>	hmm
00:22	<taupehat>	hi efudd
00:22	<taupehat>	erikh even
00:22	<taupehat>	so I found a spammer hosted at TP
00:22	<erikh>	TP?
00:22	<erikh>	for my bung hole?
00:22	<taupehat>	my mail log is full of him being rejected
00:22	<taupehat>	erikh: ThePlanet, n00b
00:22	<taupehat>	where'd opbot go?
00:22	<erikh>	eh?
00:23	<erikh>	wrong network, n00b.
00:23	<taupehat>	right
00:23	<taupehat>	I'm talking to you where you are
00:23	<taupehat>	about the network where you are not
00:23	<erikh>	grr....
00:24	<erikh>	if the system clock magically reset itself, I'm going to hurt someone..
00:24	<erikh>	this is where I start swearing
00:25	<taupehat>	hehe
00:25	<erikh>	my damn system clock reset
00:25	<erikh>	caker: can you fix the hwclock on host39?
00:35	<@mikegrb>	he did
00:36	<@mikegrb>	well, he changed the system clock
00:37	<@mikegrb>	the hardware clock isn't used by linux except for initializing its own clock at boot and as a place to store its clock on shutdown
00:37	<@mikegrb>	though not in uml guests
00:39	<taupehat>	but in xen?
00:39	<taupehat>	btw, I think I have another client lined up for you
00:39	<taupehat>	will know in a week or so
00:44	<@caker>	taupehat: cool
00:44	<@caker>	I believe it works the same way -- paravirtualized kernels use an offset from the host
00:47	[~]	taupehat nods
01:27	\|-\|	spr [~spr@c-71-195-212-252.hsd1.ma.comcast.net] has quit [Quit: spr]
01:59	<SpaceHobo>	caker: um, did host39 have trouble about 5 hours ago?
02:03	<encode>	SpaceHobo: yep it had a kernel panic
02:10	\|-\|	darkbeholder [darkbehold@c220-239-20-56.belrs4.nsw.optusnet.com.au] has quit [Read error: Connection reset by peer]
02:11	\|-\|	darkbeholder [darkbehold@c220-239-20-56.belrs4.nsw.optusnet.com.au] has joined #linode
02:31	\|-\|	andrew_j_w [~andrew@82-69-30-171.dsl.in-addr.zen.co.uk] has joined #linode
02:33	<SpaceHobo>	ouch
05:32	\|-\|	jimcooncat [~jim@pool-72-65-107-84.ptldme.east.verizon.net] has joined #linode
05:32	\|-\|	claire{H-0} [~claire_st@darkness.replica-saga.com] has quit [Read error: Connection reset by peer]
05:35	\|-\|	internat [~internat@c210-49-250-210.ipswc1.qld.optusnet.com.au] has joined #linode
05:49	\|-\|	haha [~Scripts@125.163.72.59] has joined #linode
05:57	\|-\|	JavaWoman [~Marjolein@a80-127-23-92.adsl.xs4all.nl] has joined #linode
06:04	\|-\|	haha [~Scripts@125.163.72.59] has quit []
08:06	\|-\|	internat [~internat@c210-49-250-210.ipswc1.qld.optusnet.com.au] has quit [Quit: This computer has gone to sleep]
08:39	<iggy>	caker: what panic'ed? io? net?
08:48	<@mikegrb>	ur mom
08:53	<SpaceHobo>	That old thing.
09:19	\|-\|	cooncat [~jim@lan.howeandcompany.com] has joined #linode
09:24	\|-\|	GN [~Miranda@pat.foulston.com] has joined #linode
09:35	\|-\|	sash [~sash@bas12-toronto12-1096585589.dsl.bell.ca] has joined #linode
09:35	<sash>	hi, any admin here?
09:36	\|-\|	cooncat [~jim@lan.howeandcompany.com] has quit [Quit: Leaving]
09:36	\|-\|	sash [~sash@bas12-toronto12-1096585589.dsl.bell.ca] has quit []
10:14	\|-\|	spr [~spr@c-71-195-212-252.hsd1.ut.comcast.net] has joined #linode
10:54	\|-\|	kennyparnell [~99091171@webuser.linode.com] has joined #linode
10:57	<kennyparnell>	I'm having a problem accessing my linode via ssh and http. Everything seems fine when i log in via lish though. Any ideas?
10:57	\|-\|	lele [~81c060e3@webuser.linode.com] has joined #linode
10:58	<lele>	hi all
10:58	<lele>	I'd like to speak with a linode account manager...
11:04	\|-\|	lele [~81c060e3@webuser.linode.com] has quit [Quit: CGI:IRC]
11:05	\|-\|	kennyparnell [~99091171@webuser.linode.com] has left #linode []
11:15	\|-\|	adamg [~misthos@zeus.misthos.com] has joined #linode
11:45	\|-\|	Rifkin [~rifkin@procyondesign.net] has joined #linode
12:25	\|-\|	Eman [~eman@dyn216-8-175-239.ADSL.mnsi.net] has quit [Ping timeout: 480 seconds]
12:25	\|-\|	Eman [~eman@dyn216-8-131-78.ADSL.mnsi.net] has joined #linode
12:30	<Darxus>	first google hit for ubuntu bind9 is a useful howto on linode.com, thanks
12:34	<Darxus>	is there any way I can get authoritative (used for billing) network traffic rates for my linode in a computer readable form, say total average over the last 24 hours, so I can email myself if I hit a rate that might go over my quota?
13:01	<Xel>	I think that your data is avail. in XML format somehow.
13:01	<Xel>	Caker would know exactly how
13:21	<Darxus>	that would be cool
13:23	<Darxus>	sweet: http://www.linode.com/forums/viewtopic.php?t=1558 (google searched for linode xml)
13:24	<warewolf>	NEW NEWS FRTHE FORUMS
13:24	<linbot>	New news from forums: HTTP-Retrievable Linode Stats (in XML) in Linode.com Announcements <http://www.linode.com/forums/viewtopic.php?t=1558>;
13:28	<Darxus>	how does bandwidth billing for the first month work when you're paying by the year? Do I get the full 75gigabytes during the ~11 days in this month from when I signed up?
13:29	<iggy>	everything is prorated for partial months
13:29	<iggy>	iirc
13:29	<Darxus>	k, that seems right, just don't remember seeig it said anywhere
13:47	\|-\|	adamg_mk2 [~misthos@zeus.misthos.com] has joined #linode
13:47	\|-\|	adamg [~misthos@zeus.misthos.com] has quit [Read error: Connection reset by peer]
13:47	<Darxus>	caker: http://www.linode.com/members/info/?user=[linodeUsername] shows bandwidth used in the current month, not total right?
14:38	<Darxus>	caker: can you add the current time to that? and what timezone is that year/month in?
14:39	<Darxus>	..and day, time isn't useful to me without the day of the month
14:50	\|-\|	kenny [~kenny@209.168.235.185] has joined #linode
14:52	<@caker>	Darxus: hmm .. ok -- how about under linData.Request.TimeStamp?
14:58	\|-\|	kenny [~kenny@209.168.235.185] has quit [Quit: kenny]
15:01	[~]	caker considers nolisting
15:01	<Darxus>	I don't care much what you call it or the format as long as perl module Date::Parse's str2time() can parse it
15:01	<@caker>	that's such an obvious thing -- can't believe I didn't think of it :)
15:01	<Darxus>	?
15:01	<@caker>	http://www.joreybump.com/code/howto/nolisting.html
15:03	<warewolf>	caker- that doesn't always work
15:03	<Darxus>	caker: yeah I would expect that to work, I just wouldn't expect it to stop much spam, especially since I'm already greylisting
15:03	<warewolf>	caker- spam has gotten to me through secondary MXes where it would have been blocked on primary MXes.
15:04	<warewolf>	caker- I had a store-and-forward thing going on with a friend, and his anti-spam setup wasn't even close to mine. My server was up, and the spammer got spam to me through the secondary mx.
15:04	<@caker>	I can forsee spambots being coded with a mail spool that would make grey/nolisting moot .. I mean, it's not THAT hard to add a spooler
15:04	<Darxus>	warewolf: I don't think anybody thinks nolisting will block all spam, just that it will block some at roughly no cost
15:05	<Darxus>	caker: there's a bunch of spam that gets through greylisting, they are adapting, but slower than I expected
15:05	<@caker>	yeah, anything would be better than nothing, which is my config now
15:05	<Darxus>	heh
15:08	<npmr>	caker, could i use my host as a nolisting primary mx?
15:08	<warewolf>	caker- what's your MTA?
15:08	<@caker>	warewolf: sendmail
15:08	<warewolf>	I assume postfix or exim?
15:08	<warewolf>	caker- dude.
15:08	<warewolf>	caker- you know me.
15:08	<warewolf>	caker- why havn't you pinged me on sendmail hackery.
15:08	<@caker>	npmr: you could, but if an smtp listener ever came up on the host you'd lose mail
15:09	<npmr>	caker, would an smtp listener ever come up on the host?
15:09	<warewolf>	caker- how crazy do you want your anti-spam, and is it for a company address?
15:09	<@caker>	npmr: I can't think of a reason why at the moment, but it's possible -- better to have that under your control, imo
15:10	<iggy>	you could use the null routed brazilian of the week's IP... you'd just have to change it every week
15:11	<iggy>	I should really stop making fun of brazilians
15:11	<@caker>	warewolf: yes, company addresses -- which is partly why I'm a little leary of *listing .. but I'm more comfortable with them now that I understand them better / have seen it work
15:11	<iggy>	switch back to the french or something
15:12	<warewolf>	caker- okay I'll help you later with setting you up right
15:12	<Darxus>	caker: another neat idea was to put a lower priority mx on the same machine as, but different IP from, your higher priority primary mx, and if anything goes their flagg it as spam and blacklist the sender
15:13	<@caker>	warewolf: my sendmail setup is ancient (configs, anyway) and I no longer have my .mc files I generated my configs from -- AND I've hand-hacked the sendmail.cf file over the years .. so, in other words -- it would be something I'd have to set up in a mirror of that machine; lotsawork
15:13	<warewolf>	caker- okay well you're going to need to upgrade your sendmail to the latest ver
15:13	<warewolf>	caker- and then review the hand hackery of your .cf file
15:13	<warewolf>	caker- I'll help
15:14	<warewolf>	right now I'm on the phone with IBM getting one of their appliances fixed.
15:14	<@caker>	warewolf: yeah -- that's the plan
15:14	<warewolf>	caker- bring up a linode with a clone of your info
15:15	<@caker>	warewolf: I've got it here in Parallels
15:15	<@caker>	which I set up initially just for this purpose
15:15	<@caker>	Darxus: not sure I follow that ..
15:17	<Darxus>	caker: spammers often preferentially hit backup mxes, because people tend to spam filter them less thoroughly, so the idea is to put up a backup mx that will never be up when your primary is down
15:18	<Battousai>	there's a spamassassin plugin to check for backup mx abuse
15:24	<Darxus>	caker: I see the timestamp, thanks
15:26	<@caker>	Darxus: yeah -- can't get the timezone without extra hackery. Is that good enough?
15:27	<Darxus>	caker: yup, thanks
15:27	\|-\|	adamg_mk2 [~misthos@zeus.misthos.com] has quit [Ping timeout: 480 seconds]
15:27	<Darxus>	can you tell me which timezone it is just out of curiosity?
15:27	<@caker>	eastern
15:27	<@caker>	I can add UTC offset
15:28	<Darxus>	hmm, that would probably be good
15:28	<Darxus>	because daylight savings time would screw up seconds since the beginning of the month
15:28	<@caker>	what's the typical format of that? -N UTC ?
15:29	<Darxus>	well, from an email: Date: Tue, 23 Jan 2007 16:23:25 -0500
15:29	<@caker>	ok
15:33	<Darxus>	$ ./bwalert.pl 60
15:33	<Darxus>	actual used 5.96078785974532 gB, 7.9477171463271% of 75
15:33	<Darxus>	estimation calculated from last time sample: 9.82632100582123 gB/m, 13.101761341095% of 75 gb
15:33	<Darxus>	estimation calculated from current month: 8.14395782663284 gB/m, 10.8586104355104% of 75 gb
15:36	<Darxus>	caker: you have a +, looks like it should be -
15:37	<@caker>	I was still working on it
15:37	<@caker>	<DateTimeStamp>2007-01-23 16:37:32 -0500</DateTimeStamp>
15:37	<Darxus>	k, thanks...
15:37	<@caker>	CF's GetTimeZoneInfo function is a little backwards
15:38	<@caker>	west of UTC is a positive number .. how stupid :)
15:41	<Darxus>	caker: Date::Parse parses it right, thanks
15:51	<warewolf>	the date you cut 'n pated out of an email is RFC complaint
15:51	<warewolf>	I don't remember which RFC it is though
15:56	<Darxus>	2822
15:56	<Darxus>	it's in the date man page
15:59	<Darxus>	I hate that Time::Local doesn't allow you to specify a timezone
16:02	\|-\|	iamback [~Marjolein@a80-127-23-92.adsl.xs4all.nl] has joined #linode
16:05	<@mikegrb>	caker: http://bugi.oulu.fi/~heikkiv/chan/1124311452814.jpg <-- PERSONS UKNOWNE HATH SET UP US THE CATAPULTS... ALL THINE CASTLES ARE WITHIN OUR POSSESSION, THY DOOM AWAITS THEE!
16:05	<Darxus>	caker: feel like putting the timezone offset in upSince?
16:06	<GN>	mikegrb: all your base
16:06	<warewolf>	are belong to us
16:07	<Darxus>	years ago my mom was really into cooking newsgroups, and one day she asked me what "all your flan are belong to us" meant, that was fun to explain
16:08	<GN>	I bet
16:09	\|-\|	JavaWoman [~Marjolein@a80-127-23-92.adsl.xs4all.nl] has quit [Ping timeout: 480 seconds]
16:10	\|-\|	iamback changed nick to JavaWoman
16:22	\|-\|	Netsplit helium.oftc.net <-> xenon.oftc.net quits: Redgore, flatronf700B, Beirdo
16:37	\|-\|	andrew_j_w [~andrew@82-69-30-171.dsl.in-addr.zen.co.uk] has quit [Remote host closed the connection]
16:38	\|-\|	Redgore [~Redgore@65.19.178.250] has joined #linode
16:43	<@mikegrb>	efudd: http://www.google.com/search?hl=en&lr=&q=%2812+volts%29%2F%28500milliamps%29&btnG=Search
16:43	<efudd>	Oi?
16:48	<warewolf>	grb showing off google math
16:50	<Darxus>	that stuff is great for converting between kbits/s and gigabytes/month
16:50	<Darxus>	although it uses the length of an average month
17:00	\|-\|	GN [~Miranda@pat.foulston.com] has quit [Read error: Connection reset by peer]
17:20	\|-\|	host39 [~0ca4c45a@webuser.linode.com] has joined #linode
17:20	\|-\|	fake [~fake@cpe-24-59-126-59.twcny.res.rr.com] has joined #linode
17:21	<SpaceHobo>	hurg
17:21	<@caker>	hmm
17:21	<fake>	.....
17:21	<fake>	not cool
17:21	<SpaceHobo>	:<
17:23	[~]	host39 takes a nap
17:24	\|-\|	erikh [~erikh@hollensbe.org] has quit [Ping timeout: 480 seconds]
17:24	<SpaceHobo>	host39 needs to take some medication for panic attacks
17:25	\|-\|	encode [~encode@li12-151.members.linode.com] has quit [Ping timeout: 480 seconds]
17:25	<host39>	is there a way to prevent certain db files generated by various application from becoming corrupted by a host panic? (it only happens sometimes)
17:25	<ajmitch>	host39 is dead again? how annoying
17:26	<SpaceHobo>	host39: not really. frequent dumps are your best insurance
17:26	[~]	SpaceHobo h9 berkeley DB for being unrepairable in so many situations
17:28	<fake>	no offense caker but this is getting a little too much
17:28	<linbot>	New news from forums: Reboot: host39 in System and Network Status <http://www.linode.com/forums/viewtopic.php?t=2577>;
17:28	<taupehat>	host39: you erikh?
17:29	<host39>	i've got rotated backups so maybe try to detect the error and rebuild
17:29	<host39>	negative
17:29	<taupehat>	k
17:38	<fake>	eta on back up?
17:47	<host39>	...
17:47	<fake>	what?
17:48	<host39>	we're in the same boat
17:48	<@caker>	still looking into it
17:48	<taupehat>	fake: I recall the message saying the kernel needs to be rebuilt
17:48	<ajmitch>	just relaxing, waiting to get back onto it.. :)
17:48	<fake>	ah
17:51	[~]	host39 got ping
17:52	<ajmitch>	it's a good reminder for me to setup regular backups :)
17:56	<host39>	now wait for ping from your domain... ;)
17:56	<ajmitch>	ssh would be nicer :)
18:02	<ajmitch>	yay, I'm back in
18:02	<ajmitch>	caker: thanks for getting it back up
18:06	[~]	ajmitch guesses host39 is a bit io-bound right now
18:08	\|-\|	fake [~fake@cpe-24-59-126-59.twcny.res.rr.com] has quit [Ping timeout: 480 seconds]
18:12	<host39>	yet more fun :) "Starting MySQL database server: mysqld...failed."
18:12	<ajmitch>	argh, bad day. server gone awol for work
18:16	<@caker>	host39: doesn't a quick repair tables command usually take care of that?
18:16	<host39>	yes it does
18:18	<@caker>	I've seen some distros incorporate that into the init script
18:19	<chris>	To be fair, mysqld can fail to start up for a large number of reasons
18:46	\|-\|	erikh [~erikh@hollensbe.org] has joined #linode
18:48	<Darxus>	http://www.chaosreigns.com/code/dl/linodebw.pl - it's been tested very little, I see there was another program with the same name but it seems to have vanished
18:49	<Darxus>	it reports total bandwidth used so far this month, estimated bandwidth for the month based on bandwidth used so far, estimated bandwidth for the month based on the last 3 hours (all from linode's xml), and estimated bandwidth for the month based on /proc/net/dev for the last 3 hours
18:49	<@caker>	Darxus: cool -- maybe post it on the wiki (and announce on the forums)?
18:49	<Darxus>	if anything is over the specified percentage of your quota it'll email you, including corresponding extra bandwidth costs if you're over quota
18:50	<Darxus>	caker: not till I have enough data to bother creating a gnuplot file to graph the data (with corresponding extra testing)
18:53	[~]	linbot dispenses meatloaf and au gratin potatoes
18:54	\|-\|	encode [~encode@thecoads.com] has joined #linode
18:55	<Darxus>	just added the ability to comment out the email address definition line to prevent email
18:56	<@caker>	Darxus: it would be really trick if you snarfed the username from /proc/cmdline
18:56	<@caker>	Darxus: also, that'll only be for UML Linodes (which, you can test for by the absense of umid in cmdline)
18:57	<Darxus>	is there a page anywhere with random fun stuff like this for linode?
18:58	<Darxus>	"umid" does not occur in my /proc/cmdline
18:58	<@caker>	nevermind - umid is eaten by uml itself, it's not on the cmdline ... but, you could still grab it from
18:59	<@caker>	.. the disk lines ( /dev/vgN/username-imageid or /linodes/username/imageid.fs )
18:59	<Darxus>	k, sounds good
19:01	<Darxus>	eth0=tuntap,darxus_0,fe:fd:40:47:98:28
19:01	<Darxus>	^ is that one any more or less consistant than the other two, or should I check for all...
19:02	<Darxus>	I don't have a /linodes/username/imageid.fs
19:02	<@caker>	the network stuff will display differently in Xen
19:02	<@caker>	Right -- you're on an LVM-backed host
19:04	<Darxus>	so one of those disk lines will work on oth uml and xen hosts?
19:05	<npmr>	caker, how do you set lvm volume device file ownership?
19:05	<@caker>	npmr: chown ?
19:06	<npmr>	really?
19:06	<@caker>	Yes -- am I missing something?
19:06	<npmr>	maybe i am
19:07	<npmr>	i thought there was some lvm or device mapper or udev config where you could specify ownership and permissions for those things
19:08	<@caker>	Darxus: I'm wrong again. That won't work under Xen, either. So I guess just grep for the eth device
19:08	<@caker>	I'll look for a xen-specific method
19:08	<npmr>	i tried to find a way using udev configs several months ago, but it was all so convoluted i gave up without ever figuring out whether it was even possible
19:10	<Darxus>	what's the current situation with xen on linode? I did some searching but didn't find anything
19:11	<Battousai>	i'm waiting for kvm
19:11	<Battousai>	heh
19:13	<@caker>	npmr: udev's permissions.rules? Yeah .. I gave up on that too and just reset the perms during pre-node-boot set up
19:13	\|-\|	jams [~jams@CPE-72-131-6-174.wi.res.rr.com] has quit [Read error: Connection reset by peer]
19:14	\|-\|	jams [~jams@CPE-72-131-6-174.wi.res.rr.com] has joined #linode
19:14	<@caker>	Darxus: we're done with our xen implementation, nay for a bug or two -- mainly waiting for Xen to stabilize or until something better comes along (did I mention all our new hosts are VT-enabled? :)
19:16	<Darxus>	ah, I think I saw you post that much info... like 3 years ago
19:16	<Darxus>	vt-enabled how?
19:17	<@caker>	VT capable, rather
19:20	<Darxus>	meaning what? ctrl-a,c will get me more VTs in lish?
19:20	<@caker>	VT is Vanderpool -- Intel's hardware virt support in the processoDr
19:20	<Darxus>	ohhh, that sounds way more relevant, thank you :)
19:21	<@caker>	AMD has the same thing called Pacifica (not virtual terminals) :)
19:21	<Darxus>	is vanderpool intended to be pronounced german so it sounds even more like "wonderful"?
19:22	<@caker>	ubervirt .. yeah -- I don't think so
19:22	<@caker>	!google ubervirt
19:22	<linbot>	caker: No matches found. (Search took 0.54 seconds)
19:22	[~]	caker patents
19:22	<taupehat>	Darxus: if you have a copy of "Fire in the Valley" you could probably find out the reason in the index
19:26	\|-\|	Netsplit synthon.oftc.net <-> xenon.oftc.net quits: Redgore
19:27	<ajmitch>	caker: not considering kvm or any of the new toys that have come along?
19:28	<ajmitch>	ah, virtualbox was the latest one to be announced as gpl
19:28	[~]	ajmitch has heard that full virtualisation on the recent VT-enabled chips isn't that great still
19:30	<@caker>	ajmitch: quite the opposite -- KVM seems a HELL of a lot simpler compared to Xen
19:30	<@caker>	of course, Xen works on non-vt hardware...
19:33	<ajmitch>	what are the management tools like for kvm?
19:34	<ajmitch>	since you'll want to balance I/O & CPU time still, I presume
19:41	<@caker>	I haven't messed with it yet. KVM VMs are scheduled CPU time by the host's normal scheduler; and I'm sure I'll need to hack in a disk I/O controller of some kind
19:41	<tsuyoshi>	what is the advantage of kvm or xen versus uml?
19:42	<@caker>	KVM requires hardware support; xen will use VT/Pacifica or paravirtualized/modified guest kernels on machines without hardware virtualization support, and both are faster than UML
19:42	<@caker>	although there are UML-KVM patches coming down the pipeline
19:43	<@caker>	which would essentially make UML as fast
19:48	<npmr>	does that mean linode might stick with uml long-term?
19:53	<@caker>	possibly .. I'd still like to offer hatever capabilities are of the virt tech we end up using
19:53	<@caker>	if we do KVM+UML that means we could likely do KVM+end-user custom kernels
19:57	<ajmitch>	that'd be nice
20:38	\|-\|	kenny [~kenny@c-68-58-207-118.hsd1.sc.comcast.net] has joined #linode
21:00	<Darxus>	heh, over 3 hours, the linode xml and /proc/net/dev differed by 1.7% = 0.22gB/m
21:02	<ajmitch>	there's probably some overhead somewhere
21:04	<Darxus>	caker: if you add the timezone offset before the end of the month I'm likely to properly handle the first partial billing month
21:06	\|-\|	Rifkin [~rifkin@procyondesign.net] has quit [Quit: leaving]
21:06	<Darxus>	ajmitch: shrug I'm confident linode's billing is decent and I realize those numbers are harder to get to match up than one might hope
21:07	<Darxus>	which is why I asked for the bandwidth info used for billing instead of just using my /proc/net/dev
21:09	<Darxus>	caker: I meant the timezone offset for the upSince bit, assuming that's when the account was created
21:10	<@caker>	when it was booted, and yeah -- I'll fix that one now
21:10	<Darxus>	thanks
21:11	<Darxus>	wait, you're saying if I reboot the machine that gets reset? that wouldn't be useful for handling the first month correctly then
21:11	<Darxus>	..and that's okay :)
21:12	<Darxus>	what's used to generate the very pretty bandwidth graphs? I've never managed to generate antialiased graphs
21:12	\|-\|	kenny [~kenny@c-68-58-207-118.hsd1.sc.comcast.net] has quit [Quit: kenny]
21:16	<@caker>	rrdtool
21:20	<Darxus>	ah, nice that they've made it that much prettier
21:54	\|-\|	schultmc_ [~schultmc@c-68-58-138-203.hsd1.in.comcast.net] has joined #linode
22:05	\|-\|	schultmc_ [~schultmc@c-68-58-138-203.hsd1.in.comcast.net] has quit [Quit: Client exiting]
22:09	\|-\|	myfnp [~4720e988@webuser.linode.com] has joined #linode
22:09	\|-\|	myfnp [~4720e988@webuser.linode.com] has quit []
22:10	<linbot>	New news from forums: Best CPAN mirror for Fremond Data Centre? in Linux, Apache, Mysql and PHP (LAMP) Forum <http://www.linode.com/forums/viewtopic.php?t=2578>;
22:11	\|-\|	myfnp [~4720e988@webuser.linode.com] has joined #linode
22:11	\|-\|	myfnp7 [~myfnp@71-32-233-136.slkc.qwest.net] has joined #linode
22:13	<myfnp7>	wierd returned mail from another host from a spammer domain???
22:16	\|-\|	myfnp [~4720e988@webuser.linode.com] has quit [Quit: CGI:IRC (Ping timeout)]
22:16	\|-\|	myfnp7 [~myfnp@71-32-233-136.slkc.qwest.net] has quit [Quit: myfnp7]
22:18	<efudd>	HA.
22:18	[~]	efudd looks at warewolf
22:25	<efudd>	I guess I'll update 'bw.pl' for this 'new' stuff.
22:26	<efudd>	caker, should I be changing /bw/ to /info/ in the new version?
22:26	<@caker>	efudd: yes, please
22:26	<efudd>	okee.
22:26	[~]	efudd had no idea it'd changed.
22:27	<@caker>	seemed to make more sense as info, since it's more than just bandwidth
22:27	[~]	efudd nod
22:27	<efudd>	I'm getting there are < 10 people who know/use "bandwidthSnarf"
22:28	<efudd>	... I really don't wanna rename it tho. :)
22:28	<efudd>	# bw.pl - Linode Bandwidth Utilization^W^H^Winfo Monitor via shell
22:28	<efudd>	There. that'll have to do.
22:32	<@caker>	hmm
22:32	<@caker>	# grep "BandWidth Snarf" access_log \| wc -l
22:32	<@caker>	0
22:32	<efudd>	my $useragent = "BandWidth Snarf v1.11/" . $username;
22:32	<efudd>	I just changed the version part.
22:33	<efudd>	interesting tho.
22:33	<efudd>	(grep -c btw...)
22:33	<@caker>	is $useragent being read by or fed into your UserAgent instance somehow?
22:34	<efudd>	ha
22:34	<efudd>	apparently I never did that.
22:34	[~]	efudd snickers at self
22:34	<@caker>	bummer.. :)
22:34	<efudd>	I have the focus of a gnat.
22:37	<efudd>	how about now?
22:38	<@caker>	# grep -c "BandWidth Snarf" access_log
22:38	<@caker>	2
22:38	<efudd>	ok.
22:38	[~]	efudd dumbass. :)
22:40	<efudd>	should show up as sompn like "BandWidth Snarf v1.11/mulogin"
22:40	<efudd>	s/mu/my/
22:40	<efudd>	As of Wed Jan 24 04:40:11 2007 you are currently UNDER your set limit (80%). Input Bytes Xfer: 0.84 GiBs Host: host29.linode.com Output Bytes Xfer: 2.62 GiBs Load: IDLE ------------------------ Total Bytes Xfer: 3.46 GiBs of 200.00 available.
22:40	[~]	efudd stabs IRK
22:41	<@caker>	"BandWidth Snarf v1.11/efudd"
22:41	<efudd>	aight.
22:41	<@caker>	is upSince anywhere near correct?
22:41	<efudd>	fucked if I care^Wknow. I didn't bother parsing it. ummm....
22:41	<efudd>	yeah. at least matches to the day.
22:41	<@caker>	Mine reports back a 2005 date, which doesn't even correspond to a job for my user
22:42	<efudd>	what is the definition of 'cpuConsumption' ?
22:42	<@caker>	heh .. it's bad -- it's the average CPU usage
22:42	<efudd>	Right. I'll pretend that doesn't exist then.
22:47	<@caker>	hmm
22:47	<@caker>	crap man
22:47	<efudd>	http://www.linode.com/forums/viewtopic.php?p=3179#3179
22:47	<efudd>	there. updated that original post and stuff.
22:49	\|-\|	host39_again [~0ca4c45a@webuser.linode.com] has joined #linode
22:49	<efudd>	crap?
22:49	<host39_again>	gone! ;)
22:49	<efudd>	oic.
22:50	<host39_again>	one more time and host39 holds the new record for panics per day
22:50	<@caker>	screw this. We'll ship new hardware for you guys
22:50	\|-\|	erikh [~erikh@hollensbe.org] has quit [Ping timeout: 480 seconds]
22:51	\|-\|	encode [~encode@thecoads.com] has quit [Ping timeout: 480 seconds]
22:51	<ajmitch>	down again? that box is not happy
22:52	\|-\|	host39 [~0ca4c45a@webuser.linode.com] has quit [Quit: CGI:IRC (Ping timeout)]
22:53	<ajmitch>	we'll need to put in for a migration, or will you migrate them all?
22:53	<@caker>	we'll set them all up, you can click the button
22:53	<ajmitch>	ok
22:53	<taupehat>	jeez
22:54	<taupehat>	ahh, the joys of the sysadmin never cease eh caker?
22:55	[~]	host39_again had corrupt system files both times today on host39
22:55	<host39_again>	that doesn't usually happen, is it just bad luck wrt to when it crashes?
22:58	\|-\|	VS_ChanLog [~stats@ns.theshore.net] has left #linode [Rotating Logs]
22:58	\|-\|	VS_ChanLog [~stats@ns.theshore.net] has joined #linode
23:01	<@caker>	I'll have to see what I can do about the replacement, but I can offer a switch to a new host at HE, which involves an IP switch
23:12	<host39_again>	are we migrating off host39?
23:17	\|-\|	fake [~fake@cpe-24-59-126-59.twcny.res.rr.com] has joined #linode
23:17	<fake>	is host39 dead again?
23:18	<@tasaro>	fake: yes, just restarted
23:18	<fake>	caker, can you migrate me? this is pretty unacceptable
23:19	<@caker>	fake: hardware fails -- give us a chance, please
23:20	<fake>	no worries-- just frustrated :[
23:20	<@caker>	we're in the same boat ...
23:26	[~]	host39_again gets out to push
23:26	<ajmitch>	caker: nothing available in the same DC at the moment so we can keep the ip address?
23:26	<ajmitch>	it's mainly a matter of convenience, I guess
23:27	<host39_again>	careful you don't jump into the fire, some hosts are worse
23:28	<ajmitch>	comforting to know
23:28	<host39_again>	as a matter of fact until today host39 is the most stable node of the 3 i've been on
23:29	[~]	host39_again is in no hurry to move yet ;)
23:30	<@tasaro>	It would be great to get whoever doesn't mind migrated to a new host at HE, but we're not going to force anyone to change IPs
23:32	<fake>	heh i got migrated out of HE a few months back
23:32	\|-\|	Newsome [~sorenson@adsl-75-0-136-70.dsl.chcgil.sbcglobal.net] has joined #linode
23:32	[~]	ajmitch doesn't mind, it just means fixing up dns for a number of domains
23:33	[~]	host39_again was migrated there by accident once, had to move back ;)
23:34	[~]	host39_again can't change ip without clocking a few billable hours
23:35	<ajmitch>	tasaro: new hardware at HE, is it?
23:36	<host39_again>	if there's a hardware problem on host39 will it keep crashing if we don't move?
23:37	[~]	ajmitch would think so
23:37	<@tasaro>	ajmitch: Yes, new hardware
23:37	<ajmitch>	tasaro: I think I can handle that :)
23:38	<@tasaro>	host39_again: We're still looking into the cause, but I think it's safe to say we will probably migrate everyone off host39
23:38	<@tasaro>	Either another TP host or a new HE host
23:38	<ajmitch>	will the new hosts be using xen yet, or still uml?
23:38	<host39_again>	thanks tasaro, good to know.
23:38	[~]	ajmitch presumes uml for everyone still
23:39	<@caker>	ajmitch: they're set up with UML, but are just a reboot away from Xen
23:39	\|-\|	host39_again changed nick to sol
23:40	<npmr>	what kind of plan is on host39?
23:41	<@caker>	384/512
23:41	<npmr>	ok
23:41	<npmr>	oh, duh
23:41	<npmr>	i knew that
23:41	<@caker>	heh
23:41	<npmr>	frotz is a 512
23:41	<@caker>	right
23:44	[~]	ajmitch would love to see how some of the live migration with xen would go
23:45	<ajmitch>	except that it needs shared storage
23:45	<sol>	i'm curious what hardware failed?
23:45	<fake>	whoa i have 640 ram
23:45	<fake>	crazy
23:45	[~]	sol has 800 =)
23:45	<ajmitch>	lucky you :)
23:46	<npmr>	i had 1200, briefly... in august
23:46	[~]	sol still wants more (and thanks linode for the upgrade)
23:46	[~]	ajmitch still only has 384
23:47	<sol>	1200 sounds about right but host39 had 0 extra =(
23:47	[~]	Eman has 64
23:47	<npmr>	mine was due to a typo in the host's config file
23:47	<npmr>	(hee hee @ mikegrb)
23:48	<sol>	sweet
23:48	<@caker>	http://www.theshore.net/~caker/pics/Linode/linode159744.jpg <-- Linode 159744
23:48	<npmr>	Eman, 487 days uptime?
23:48	<ajmitch>	hehe nice
23:48	<Eman>	2 years + couple of days
23:48	[~]	ajmitch wonders how we get upgrade goodies
23:49	<ajmitch>	nicest box I've had to install linux on had 16GB RAM, 8 cores, 4 threads per core
23:49	[~]	Eman <3 host9
23:49	<ajmitch>	it wouldn't be so bad for running stuff on :)
23:49	<sol>	2 years uptime on linode?
23:50	<Eman>	yes
23:50	<npmr>	yeah, a friend of mine has a linode on host 9
23:50	<@caker>	setterm -blank 0 really should be the default, or at least wakes the monitor on BUGs and OOPs
23:50	<npmr>	looks like he rebooted a couple of months ago
23:50	<ajmitch>	a decent serial console would be good
23:50	<@caker>	ajmitch: we have them on newer boxen -- I didn't matter this last case -I left remote console attached
23:50	[~]	sol must be on the bleeding edge, never saw more than 60 days but 30 is more common
23:51	<@caker>	no output
23:51	<ajmitch>	how annoying
23:51	<npmr>	i had a server up for almost a year once
23:51	<@caker>	sol: the vast majority of hosts are rock solid
23:51	<npmr>	i booted it up when i moved in
23:51	<npmr>	shut it down a couple of days before i moved out
23:51	<taupehat>	heh
23:51	<npmr>	it was a 12-month lease
23:52	<taupehat>	npmr: hold that thought
23:52	<sol>	i can accept that i'm having bad luck, one host had a drive failure, etc
23:52	<taupehat>	gonna show you something
23:52	<@caker>	sol: you and a few others have had terrible luck with hosts
23:52	<taupehat>	friend of mine has an old alpha
23:53	<ajmitch>	sol: I was setting up a box in a local hosting place, it was going down every week or two with hardware issues
23:53	<ajmitch>	corrupting files
23:53	<ajmitch>	now that was annoying
23:53	<sol>	sounds like host 53 a few months back ;)
23:53	\|-\|	taupehat_ [gloin@jeancharles.org] has joined #linode
23:53	<taupehat_>	ahh, here we go
23:53	<sol>	(iirc)
23:53	<fake>	dear god stay up long enough for this download to finish
23:53	<taupehat_>	in all my non-anonymous glory =[
23:53	<taupehat_>	9:54PM up 1001 days, 7:30, 1 user, load averages: 0.00, 0.00, 0.00
23:54	<Eman>	that uptime is pure win
23:54	<ajmitch>	taupehat_: impressive
23:54	<taupehat_>	aye
23:54	<taupehat_>	it's been moved twice
23:54	<taupehat_>	with daisy-chained UPSs to make sure it stayed up
23:54	<ajmitch>	haha
23:55	<sol>	you can only hope i don't migrate to your host, hehe =(
23:55	<npmr>	i think it was alan cox who had some old system with a linux 0.999 kernel still running in about 1996
23:55	<taupehat_>	hehe
23:55	<npmr>	that thing had the highest pc uptime i've ever heard of
23:55	<npmr>	it was over 1200 days
23:55	<taupehat_>	that'd be an AC box allright
23:55	<taupehat_>	this thing's up there
23:55	<npmr>	it's also not a pc
23:56	<taupehat_>	nope
23:56	<taupehat_>	it's pure d\|i\|g\|i\|t\|a\|l goodness
23:56	<@caker>	host9
23:56	<Eman>	haha
23:56	<npmr>	alpha ftw
23:56	<@caker>	00:56:20 up 995 days, 9:01, 400 users, load average: 1.19, 1.04, 0.92
23:56	<taupehat_>	caker: rad!
23:56	<@caker>	ignore the num users -- utmp bug
23:56	<npmr>	:)
23:56	<@caker>	there are a few others like that
23:56	<taupehat_>	where?
23:56	<taupehat_>	those linodes?
23:56	<npmr>	taupehat, that's host9
23:56	<taupehat_>	no joke?
23:56	<sol>	put me on host9 let's see what happens
23:57	<taupehat_>	heh
23:57	<npmr>	ha ha
23:57	<@tasaro>	ha
23:57	<taupehat_>	sol: don't feel too bad
23:57	<@caker>	host21 is the same; the rest have since had kernel upgrades
23:57	<taupehat_>	I was on a notorious host for a while
23:57	<Eman>	51?
23:57	<taupehat_>	then I volunteered to go on a xennode
23:57	<ajmitch>	how stable was the xen node?
23:57	<taupehat_>	obviously, I don't freak out if the servere bounces around a bit
23:57	<npmr>	the original host4 had panic attacks like nothing else
23:58	<taupehat_>	ajmitch: at first it was pretty wild
23:58	<taupehat_>	but it's been pretty damn good lately
23:58	<npmr>	iirc, host4 was the first host to ever be replaced
23:58	<@caker>	npmr: bios update fixed it
23:58	<npmr>	ah
23:58	<taupehat_>	hmm
23:58	[~]	ajmitch has had some impressive kernel panics & odd stuff with ubuntu+xen, but that's mainly due to running development kernels straight from the developer's crack pipe
23:58	<@caker>	due to a problem with the e1000 stuffs
23:58	[~]	taupehat_ is drinking chardonnay from the bottle
23:59	[~]	npmr is drinking pinot from the barrel
23:59	<ajmitch>	caker: will migration be available sometime tonight or so?
23:59	<npmr>	(in the future)
23:59	<ajmitch>	(well, 7pm here, so the night is young)
23:59	<taupehat_>	heh
23:59	<taupehat_>	ajmitch: you in HI?
23:59	<ajmitch>	NZ
---	Log	closed Wed Jan 24 00:00:00 2007