TheGrebs.com:IRC Logs:#linode:2007:01:31

---	Log	opened Wed Jan 31 00:00:25 2007
00:03	\|-\|	myfnp [~myfnp@70-56-103-133.slkc.qwest.net] has quit [Quit: myfnp]
00:03	\|-\|	brc [bruce@i.am.someasshole.com] has joined #linode
00:18	\|-\|	ThE__CroW_ [demonicpag@ip68-101-0-125.pn.at.cox.net] has joined #linode
00:20	\|-\|	dc0e [~dc0e@c-69-243-116-179.hsd1.md.comcast.net] has quit [Quit: ↻ or ↓]
00:28	<ThE__CroW_>	how are additional ips added in a gentoo distro?
00:40	\|-\|	dc0e [~dc0e@c-69-243-116-179.hsd1.md.comcast.net] has joined #linode
01:16	<ThE__CroW_>	how are additional ips added in a gentoo distro?
01:32	<encode>	by little green men
01:39	<Battousai>	see /etc/conf.d/net.example
01:40	\|-\|	f8_ [~buddyw@f8.budw.net] has quit [Server closed connection]
01:40	\|-\|	f8 [~buddyw@f8.budw.net] has joined #linode
01:47	\|-\|	fake [~fake@cpe-24-59-126-59.twcny.res.rr.com] has quit [Ping timeout: 480 seconds]
02:05	\|-\|	rick [Ap0ll0@modemcable129.87-80-70.mc.videotron.ca] has quit [Server closed connection]
02:05	\|-\|	rick [Ap0ll0@modemcable129.87-80-70.mc.videotron.ca] has joined #linode
02:28	\|-\|	andrew_j_w [~andrew@82-69-30-171.dsl.in-addr.zen.co.uk] has joined #linode
02:53	\|-\|	kvandivo [~kvandivo@ppp-70-225-162-235.dsl.chmpil.ameritech.net] has quit [Server closed connection]
02:53	\|-\|	kvandivo [~kvandivo@ppp-70-225-162-235.dsl.chmpil.ameritech.net] has joined #linode
02:54	\|-\|	linbot [~supybot@ns.theshore.net] has quit [Server closed connection]
02:55	\|-\|	linbot [~supybot@ns.theshore.net] has joined #linode
03:01	\|-\|	AndyHat [~andyhat@user-10lf9kj.cable.mindspring.com] has quit [Ping timeout: 480 seconds]
03:01	\|-\|	AndyHat [~andyhat@user-10lf9kj.cable.mindspring.com] has joined #linode
03:03	\|-\|	smiffy [~cb1c9fa8@webuser.linode.com] has joined #linode
03:04	<smiffy>	To mis-quote Paul Atreides, "He is gone". I just killed my colo and issued the final shutdown - the future is now entirely in the hands of my ghola Linodes.
03:26	\|-\|	flatronf700B [~flatronf7@202.75.186.154] has quit [Server closed connection]
03:34	<smiffy>	Anyone here installed Slashem on their Linodes?
03:38	\|-\|	npmr [~inkblot@dorothy.movealong.org] has quit [Server closed connection]
03:38	\|-\|	npmr [~inkblot@dorothy.movealong.org] has joined #linode
03:38	<npmr>	linbot, roulette
03:38	<linbot>	npmr: click
03:43	<smiffy>	Hmm.
03:44	<smiffy>	linbot, roulette
03:44	<linbot>	BANG Hey, who put a blank in here?!
03:44	[~]	linbot reloads and spins the chambers.
03:44	[~]	smiffy reels from the bang.
03:49	<smiffy>	linbot, roulette
03:49	<linbot>	smiffy: click
03:50	<smiffy>	Ah, I live another day.
03:51	\|-\|	Redgore [~Redgore@65.19.178.250] has quit [Server closed connection]
03:52	\|-\|	Redgore [~Redgore@65.19.178.250] has joined #linode
03:55	\|-\|	smiffy [~cb1c9fa8@webuser.linode.com] has quit [Quit: CGI:IRC]
03:57	\|-\|	encode [~encode@thecoads.com] has quit [Server closed connection]
03:57	\|-\|	encode [~encode@thecoads.com] has joined #linode
03:58	\|-\|	Dreamr_3 [~Dreamer3@74-129-108-228.dhcp.insightbb.com] has quit [Quit: Leaving]
04:25	\|-\|	Beirdo [~gjhurlbu@beirdo.usercloak.oftc.net] has quit [Server closed connection]
04:25	\|-\|	Beirdo [~gjhurlbu@beirdo.usercloak.oftc.net] has joined #linode
04:31	\|-\|	smiffy [~cb1c9fa7@webuser.linode.com] has joined #linode
04:37	\|-\|	smiffy [~cb1c9fa7@webuser.linode.com] has quit [Quit: CGI:IRC (Ping timeout)]
05:00	\|-\|	taupehat [me@taupehat.com] has quit [Server closed connection]
05:01	\|-\|	taupehat [me@taupehat.com] has joined #linode
05:50	\|-\|	internat [~internat@c210-49-250-210.ipswc1.qld.optusnet.com.au] has joined #linode
05:51	\|-\|	Internat-bed [biteme@c210-49-250-210.ipswc1.qld.optusnet.com.au] has quit []
08:20	\|-\|	VSM [~4170683e@webuser.linode.com] has joined #linode
08:21	<VSM>	Anyone know how to get a hold of Chris Aker?
08:21	[~]	Beirdo points to caker
08:24	<VSM>	Thanks... he doesn'
08:24	<VSM>	t respond
08:24	<SpaceHobo>	14:24 [Linode] CTCP TIME reply from caker: Wed Jan 31 09:24:38 2007
08:25	<@mikegrb>	mmm cake
08:25	<SpaceHobo>	it's still early morning in the land of cake
08:25	<Beirdo>	ah, well, if you are a customer already, putting in a ticket might help get your issue resolved
08:25	<SpaceHobo>	yeah
08:25	<SpaceHobo>	filing a ticket is a great way to get attention
08:25	<VSM>	Not really and issue... we are trying to do an article about him
08:26	<Beirdo>	ah.
08:26	<VSM>	(Virtual-strategy.com
08:26	<Beirdo>	you would likely have to wait until he's up and responsive, I guess :)
08:26	<VSM>	And we are also a customer
08:27	<SpaceHobo>	haha
08:27	<VSM>	We are doing an article about UML with Jeff Dike and thought Chris would be a good success story
08:28	<Beirdo>	cool
08:28	<VSM>	I'll try him later...
08:28	<VSM>	thanks
08:30	\|-\|	Kurt [1000@evvlinlwt-nas-07-s59.cinergycom.net] has joined #linode
08:34	<heidi>	actually that time is wrong
08:34	<SpaceHobo>	Oh.
08:34	<heidi>	it is 8:30 for caker
08:34	<heidi>	so even earlier
08:34	<SpaceHobo>	14:34 [Linode] CTCP TIME reply from heidi: Wed Jan 31 08:34:37 2007
08:35	<heidi>	hmm
08:35	<heidi>	yeah, I know he is in central time as well
08:35	<heidi>	nashville is pretty close to the border
08:35	<heidi>	i wonder why his shows up that way
08:36	<heidi>	http://www.timeanddate.com/worldclock/results.html?query=nashville
08:36	<heidi>	that gives the acurate time to where he is at
08:37	\|-\|	linville [~linville@nat-pool-rdu.redhat.com] has joined #linode
08:40	\|-\|	Kurt [1000@evvlinlwt-nas-07-s59.cinergycom.net] has quit [Ping timeout: 480 seconds]
09:24	\|-\|	VSM [~4170683e@webuser.linode.com] has quit [Quit: CGI:IRC (Ping timeout)]
09:30	\|-\|	internat [~internat@c210-49-250-210.ipswc1.qld.optusnet.com.au] has quit [Quit: This computer has gone to sleep]
10:13	\|-\|	VSM [~4170683e@webuser.linode.com] has joined #linode
10:16	\|-\|	GN [~Miranda@pat.foulston.com] has joined #linode
10:25	\|-\|	fake [~fake@cpe-24-59-126-59.twcny.res.rr.com] has joined #linode
10:26	\|-\|	VSM [~4170683e@webuser.linode.com] has quit [Quit: CGI:IRC (Ping timeout)]
10:34	\|-\|	ThE__CroW_ [demonicpag@ip68-101-0-125.pn.at.cox.net] has quit [Read error: Connection reset by peer]
10:35	\|-\|	ThE__CroW_ [demonicpag@ip68-101-0-125.pn.at.cox.net] has joined #linode
10:52	\|-\|	dave_ [~chatzilla@64.157.152.10] has joined #linode
10:53	<dave_>	hello... can anybody tell me (1) where are linode's vpses geographically located? (2) a Test IP I can ping to determine my latency to them?
10:57	<Darxus>	dave_: panic.chaosreigns.com is my linode, you're welcome to ping it
10:57	<Darxus>	there are two geographical locations
10:57	<fake>	fremont, ca
10:57	<iggy>	um... so when did I get moved to host71?
10:57	<fake>	and somewhere in texas
10:57	<iggy>	dallas
10:58	<iggy>	theplanet in dallas and HE in fremont
10:58	<dave_>	hmm, ok cool. where do new vpses go, in dallas?
10:59	<npmr>	whereever there's space
10:59	<iggy>	wherever there's space available
10:59	<Darxus>	dave_: I believe you can pick
10:59	<dave_>	heh, ok
10:59	<npmr>	you can request a data center
11:00	<Darxus>	dave_: if it is not obvious, the people in this channel who have ops are the ones with the authoritative information, and are not talking right now :)
11:00	<npmr>	others have been here for years and just know everything
11:01	<npmr>	so the two data centers are ThePlanet Internet Services of dallas, tx and Hurricane Electric of fremont, ca
11:01	<Darxus>	I feel like the topic should have a disclaimer along the lines of what I just said :)
11:01	<npmr>	for reference, if you need something to ping or traceroute, host1.linode.com is in dallas and host9.linode.com is in fremont
11:01	<iggy>	"The Linode Virtual Server hang out"
11:02	<iggy>	it doesn't get any more un-officialer than that
11:02	<Darxus>	iggy: and official support channel :)
11:02	<dave_>	npmr: thanks
11:02	<iggy>	Darxus: I was quoting the /topic
11:02	<npmr>	if there's available space for the plan you want at both data centers, they'll accomodate a request for a specific one
11:02	<dave_>	dallas is doing well, pinging at 40ms, but fremont pings at nearly 110ms to me in nyc
11:03	<npmr>	ah, yeah
11:03	<npmr>	that's the speed of light talking
11:03	<fake>	yeah
11:03	<fake>	cross country RTT's are on the order of 100ms
11:03	<npmr>	minimum
11:03	<npmr>	probably a little over that, minimum, actually
11:04	<Darxus>	so when are they putting up machines in the northeast? :)
11:04	<iggy>	I heard talk of atlanta at one point
11:04	<dave_>	yeah, I'd definitely want the dallas... or better yet, new jersey :)
11:04	<npmr>	dave_, if those 80ms are that important, you may be happy to know that caker is looking at a third data center
11:04	<fake>	so 110 is actually good
11:04	<Darxus>	npmr: where?
11:04	<npmr>	i believe it's in atlanta, but there is no official word yet
11:04	<dave_>	npmr: do you know if it's weekly or daily backups of the vpses?
11:04	<iggy>	no backups
11:04	<npmr>	dave_, there are no backups
11:05	<npmr>	the hosts have raid1
11:05	<fake>	hah zing
11:05	<npmr>	backups of your linode would be up to you
11:05	<dave_>	npmr: nothing offsite?
11:05	<Darxus>	I run a cron job that rsyncs all of my linode down to my workstation at home every night
11:05	<Darxus>	(which is hosted over cablemodem)
11:05	<Darxus>	works great
11:05	<iggy>	I use unison, but there's nothing historical there
11:05	<npmr>	dave_, no backups at all
11:06	<npmr>	onsite or offsite
11:06	<Darxus>	rsync --link-dest=/home/darxus/bak/panic-`date -d yesterday +\%F`
11:06	<Darxus>	^ rocks
11:06	<npmr>	personally, i back up my linode to a box at home
11:06	<dave_>	I suppose dallas isn't really a target. nobody messes with texas
11:06	<iggy>	damn right they don't
11:06	<iggy>	!
11:06	<npmr>	dave_, fwiw, linode has only ever lost both drives on a host once in three years
11:06	<npmr>	that was maybe a month or two ago
11:07	<Darxus>	how many hosts are there?
11:07	<dave_>	npmr: hmm ok. not entirely inspiring but thanks for the tip
11:07	<iggy>	I'm apparently on host71 now, so at least that many
11:08	<npmr>	PING host71.linode.com (64.71.152.11): 56 data bytes
11:08	<npmr>	64 bytes from 64.71.152.11: icmp_seq=0 ttl=51 time=64.7 ms
11:08	<iggy>	dave_: that's what you get for 19.95 a month
11:08	<Darxus>	iggy: you sound far more confident than I that they haven't skipped any
11:08	<npmr>	ping: unknown host host72.linode.com
11:08	<npmr>	dave_, linode is totally diy
11:08	<iggy>	Darxus: they don't.. in fact they go in a fill in the gaps when one is decommissioned
11:08	<npmr>	dave_, they host the server
11:08	<Darxus>	iggy: neat
11:08	<npmr>	dave_, everything else is up to you
11:09	<Darxus>	yeah, it's just like having your own machine in a colo, except if the hardware brakes somebody else deals with it
11:09	<dave_>	npmr: yeah, I can imagine. I've been using johncompanies for the past 3 years which is similar. They have good support and take care of the occassional DNS record update, that's about it.
11:10	<iggy>	linode has dns manager, which I've been playing with and loving recently
11:10	<fake>	linode has more tools than johncompanies
11:10	<fake>	from what i remember when i switched
11:10	<npmr>	Darxus, other than host56, which is the xen beta test server, hosts 1-71 are all online and functioning uml hosts
11:10	<fake>	you can do almostanything yourself.
11:10	<Darxus>	almost? :)
11:11	<fake>	:]
11:11	<iggy>	and you always have the cheery faces in here to help you through the hard stuff
11:11	<npmr>	totally
11:11	<dave_>	iggy: that's cool
11:11	<npmr>	except for efudd
11:11	<npmr>	he knows everything, but he's not cheery
11:11	<dave_>	hehe
11:12	<Darxus>	I just signed up a couple weeks ago, I'm paying the same amount as I was for colo for more bandwidth, same amount of disk space, and I'm nolonger responsible for the hardware... and I have even more control over my virtual machine than I did my phisical machine - remote reboots, console access via ssh... loving it
11:12	<npmr>	until fairly recently, i was considering upgrading from a linode to a colo'd box
11:13	<dave_>	is the linode ram limits burstable at all?
11:13	<Darxus>	npmr: why?
11:13	<npmr>	but linode keeps upgrading the plans and grandfathering in all the existing customers
11:13	<npmr>	Darxus, i was starting to outgrow the resources
11:13	<Darxus>	dave_: no, your virtual machine gets assigned a specific amount of physical ram
11:13	<dave_>	I
11:13	<Darxus>	bandwidth, cpu, and disk I/O are burstable, ram and disk space are not
11:14	<npmr>	disk i/o is often a sticking point
11:14	<Darxus>	npmr: oh?
11:14	<dave_>	Darxus: ah, ok, that's too bad. I'd probably start hitting ooms sooner than.
11:14	<npmr>	the i/o limiter is not mentioned anywhere on the signup page
11:14	<npmr>	and most people are unaccustomed to working within a disk i/o quota
11:15	<Darxus>	npmr: I'd rather have it than have someone I'm sharing a machine with using all the disk I/O...
11:15	<Darxus>	ah
11:15	<Darxus>	seems like an obvious enough limit to me
11:15	<Darxus>	dave_: ooms?
11:15	<npmr>	Darxus, out of memory
11:15	<Darxus>	ah
11:15	<dave_>	Darxus: thrown by the java jvm
11:16	<npmr>	ah, yeah
11:16	<Darxus>	ahh, yeah I'm not running java
11:16	<Darxus>	and I should move the couple things I have using mysql off of it, because it would be easy
11:16	<npmr>	heavy swapping is usually what triggers people's problems with the i/o limiter
11:17	<npmr>	but really any i/o uses tokens
11:18	<Darxus>	http://www.linode.com/wiki/index.php/IO_Tokens
11:18	<Darxus>	I'm using 2.4% of the token fill rate
11:19	<npmr>	see also, http://ratemylinode.com/
11:19	<npmr>	it's run by a linode customer
11:19	<npmr>	there are about fifty other linode customers all volunteering their i/o token usage stats and they're all compiled on the rml site
11:20	<npmr>	Darxus, i think your usage is pretty typical, but higher usage is not rare
11:20	<Darxus>	whoa there's a recommended mysql config?
11:21	<npmr>	yes there is
11:21	<npmr>	iirc, there are also recommended configs for php and apache as well
11:22	<npmr>	at least in terms of memory-related parameters
11:22	<npmr>	ha ha, nova's i/o quota got docked!
11:22	<npmr>	caker, is nova supposed to be penalized for high i/o?
11:23	<Darxus>	what's nova? host of the linode website?
11:24	<npmr>	nova is www.linode.com
11:24	<npmr>	it runs on its own host
11:24	<Darxus>	how does ratemylinode.com pull the io_status?
11:25	<npmr>	there is a perl script that participating linodes run every five minutes
11:25	<Darxus>	ah, cool
11:29	<npmr>	!weather 60625
11:29	<linbot>	npmr: Temperature: 13.8°F / -10.1°C \| Humidity: 56% \| Pressure: 30.16in / 1021.2hPa \| Conditions: Overcast \| Wind Direction: West \| Wind Speed: 5.0mph / 8.0km/h; This Afternoon - Becoming mostly cloudy. Highs 15 to 20. Lowest southwest winds 10 to 20 mph.; Tonight - Mostly cloudy with a 50 percent chance of light snow. Not as cold. Lows 14 to 18. Wind chills as low as zero to 5 above zero. Southwest winds 10 to 20 (1 more message)
11:29	<npmr>	we finally have a normal winter
11:29	<npmr>	!more
11:29	<linbot>	npmr: mph.; Thursday - Mostly cloudy. A chance of flurries. Little or no snow accumulation. Highs in the upper 20s. West winds 10 to 15 mph.;
11:52	\|-\|	adamg [~misthos@zeus.misthos.com] has joined #linode
12:00	\|-\|	afv-13 [~afv@rkdp-ip-nas-1-p453.telkom-ipnet.co.za] has joined #linode
12:06	<@tasaro>	iggy: http://www.linode.com/forums/viewtopic.php?t=2591
12:21	<afv-13>	tasaro: i take it that's one of the older hosts?
12:26	<@tasaro>	afv-13: host58 is not that old
13:17	\|-\|	VSM [~4170683e@webuser.linode.com] has joined #linode
13:22	<Beirdo>	!weather 00646
13:22	<linbot>	Beirdo: Temperature: 84.7°F / 29.3°C \| Humidity: 46% \| Pressure: 29.94in / 1013.8hPa \| Conditions: Scattered Clouds \| Wind Direction: SSE \| Wind Speed: 5.4mph / 8.7km/h; Today - Mostly sunny early in the morning then becoming partly cloudy. Isolated showers in the late morning and afternoon. Highs around 83. East winds 5 to 15 mph. Chance of rain 20 percent.; Tonight - Partly cloudy. Isolated showers early in the (1 more message)
13:22	<Beirdo>	!more
13:22	<linbot>	Beirdo: evening. Lows 75 lower elevations ranging to 61 higher elevations. Southeast winds 5 to 10 mph. Chance of rain 20 percent.; Thursday - Mostly sunny in the morning...then partly cloudy with isolated showers in the afternoon. Highs around 83. Northeast winds 5 to 15 mph. Chance of rain 20 percent.;
13:23	<Beirdo>	it'll do
13:28	<@caker>	!cnn 00646
13:29	<@caker>	ahh, Puerto Rico
13:32	<iggy>	so 58 was in pretty bad shape I guess
13:39	<@tasaro>	58 should be good to go after we send new drives
13:40	<npmr>	was it just a failed raid component?
13:41	\|-\|	Beezlebub [LinodeJava@host86-137-41-253.range86-137.btcentralplus.com] has joined #linode
13:44	\|-\|	Beezlebub [LinodeJava@host86-137-41-253.range86-137.btcentralplus.com] has quit []
13:47	\|-\|	Rifkin [~rifkin@procyondesign.net] has joined #linode
14:06	<@caker>	npmr: yeah
14:07	<npmr>	how often is that happening now?
14:07	<@caker>	first one of the Hitachi drives
14:09	<Beirdo>	yes, Puerto Rico. :
14:09	<Beirdo>	:) even (stupid keyboard)
14:11	<npmr>	caker, no, i mean across the whole host pool, how often are drives needing replacement these days?
14:11	[~]	npmr just had a weird series of unrelated drive failures at the office
14:12	<@caker>	npmr: that seems to be the theme of my nights, these past few weeks
14:12	<@caker>	although in nova'
14:12	<@caker>	nova's case it was a PSU (or something worse)
14:13	<npmr>	(solar flares?)
14:14	<npmr>	a couple of weeks ago at the office, we replaced failed drives in three out of five raid1 arrays
14:14	<npmr>	these drives were only very roughly the same age, but they did all have 100% duty-cycle
14:22	<warewolf>	sda: number of hours powered up = 36804.10
14:22	<warewolf>	sdb: number of hours powered up = 41340.65
14:22	<warewolf>	sdc: number of hours powered up = 14293.57
14:23	<warewolf>	sdd: uh .. ?
14:23	<warewolf>	sde: number of hours powered up = 14275.82
14:23	<@caker>	warewolf: what's that from? smartctl/smartd or something?
14:23	<warewolf>	hda: 9 Power_On_Hours 0x0012 036 036 000 Old_age Always - 47047
14:23	<warewolf>	caker- yeah
14:23	<warewolf>	hdb: 9 Power_On_Hours 0x0032 044 044 000 Old_age Always - 41201
14:24	<warewolf>	hdc: 9 Power_On_Hours 0x0032 075 075 000 Old_age Always - 18930
14:24	<warewolf>	and that's it for my primary file server
14:24	<warewolf>	what's the oldest
14:24	<warewolf>	the 47k one okay
14:24	<warewolf>	15:23 < purl:#perl> 47047 hours is 5.3671 years.
14:26	<warewolf>	wow, WDC WD205BA 5 years spinning time.
14:26	<warewolf>	westen digital++
14:27	<npmr>	i recently replaced a drive (not due to failure) with over 52000 hours on it
14:27	<npmr>	i believe it was also a wd
14:27	<Battousai>	thats good to know
14:27	<Battousai>	i got a wd getting up there
14:28	<npmr>	also, less than 100 spin-ups over the life of the drive :)
14:28	<@caker>	we switched to Seagate Barracuda ES "Enterprise" drives for the last 25 hosts or so
14:28	[~]	caker is so sick of drive problems
14:30	<npmr>	i went with "enterprise" drives with all of my recent replacements too
14:30	<npmr>	1.2 million hours mbtf according to the manufacturer
14:36	<Darxus>	caker: do you try to avoid raid mirrors composed of identical drives?
14:36	<@caker>	Darxus: no?
14:36	<Darxus>	caker: you're supposed to, to reduce the likelyhood of simultaneous fialures
14:37	<@caker>	drives from different manufacturers?
14:37	<@caker>	what a pain :)
14:38	<Darxus>	hah
14:38	<Darxus>	looking up the document that suggested it...
14:38	<Darxus>	there is apparently a not-insignificant possibility of disks from the same batch failing simultaneously
14:38	<Darxus>	like due to the same power dip
14:38	[~]	caker wants sugarcube sized quantum storage devices NOW
14:41	<dave_>	caker: 2009
14:47	<warewolf>	Darxus: no, you just buy extras, and have hot spares in the chassis when one drive fails.
14:48	<warewolf>	Darxus: having different manufacturer drives will generally lead to poor performance, different seek times, different cache sizes, different DISK size, etc.
14:48	<warewolf>	Darxus: you could end up wasting space on larger drives
14:50	<Darxus>	warewolf: hot spares don't help if all drives in the array decide to get sensitive to the same minor power fluctuation
14:56	\|-\|	VSM [~4170683e@webuser.linode.com] has quit [Quit: CGI:IRC]
14:56	<@caker>	^-- neat stuff
14:57	<@caker>	Darxus: I'm going to second what warewolf just said re different disk manus in the same raid
14:58	<Darxus>	heh
14:58	<@caker>	# ping host76.linode.com
14:58	<@caker>	PING host76.linode.com (64.71.152.16) from 67.18.92.50 : 56(84) bytes of data.
14:58	<@caker>	64 bytes from host76.fremont.linode.com (64.71.152.16): icmp_seq=0 ttl=51 time=64.992 msec
14:58	<@caker>	go tasaro!!
14:58	<warewolf>	Darxus: then you need power conditioning :P
14:59	<Darxus>	for the record, the raid 1 in my colo'd box has identical drives :P
14:59	<warewolf>	you can't get away with _trying_ to play enterprise class redundancy without going full tilt.
14:59	\|-\|	Kurt [1000@evvlinlwt-nas-07-s448.cinergycom.net] has joined #linode
15:00	<encode>	an interesting thing about disk failures - we have a number of old intel rackmoutn boxes here at work, and the disks fail relatively regularly - maybe one failure every month or two across all the servers
15:00	<encode>	interestingly we had a box that failed much more regularly
15:00	<warewolf>	Darxus: why are you standing up straw men, and then knocking them self over yourself?
15:00	<encode>	put an expensive ibm raid controller in it
15:00	<encode>	that we just had lying around as a spare for another box
15:00	<encode>	and left the same disks in it
15:00	<encode>	and its been going fine ever since
15:01	<@caker>	encode: more cache / less stress on the disks?
15:01	<encode>	so disk failures aren't necessarily the fault of the disks themselves
15:01	<@caker>	yeah, that's pretty interesting
15:01	<encode>	caker: i think better utilisation of the disk - maybe cos it has more cache
15:01	<Darxus>	warewolf: I believe it's worth using multiple vendors in any raid array on which multiple production hosts depend
15:01	<Darxus>	my server is just personal, not important to anybody
15:01	<encode>	but im inclined to believe its more advanced programming
15:01	<warewolf>	Darxus: quit arguging both sides of the field, you're confusing.
15:01	<@caker>	encode: ok -- less trashing / better coalescence / etc
15:02	<@caker>	*thrashing
15:02	<encode>	yeah
15:02	<warewolf>	cache too, probally.
15:02	<encode>	of course, it could be a one off case, complete coincidence
15:03	<npmr>	i ordered replacement drives once, exact same manufacturer and model as the existing ones
15:03	\|-\|	marc_in_lux [~53d99551@webuser.linode.com] has joined #linode
15:03	<npmr>	and they were 36KB smaller
15:03	<npmr>	and the raid wouldn't take them
15:03	<warewolf>	npmr: scsi or ide?
15:03	<marc_in_lux>	good evening
15:03	<npmr>	scsi
15:04	<encode>	thats bizarre
15:04	<warewolf>	you can fix that
15:04	<warewolf>	the manufacturer reserved more blocks for bad block remapping
15:04	<marc_in_lux>	just wanted to mention that network graphs don't work anymore, host 56
15:04	<warewolf>	there are scsi utils that can let you adjust those paramaters of drives
15:04	<npmr>	interesting
15:04	<npmr>	scsi utils for linux?
15:05	<warewolf>	the last one I used was scsiutil.exe :/ sadly.
15:05	<@caker>	marc_in_lux: yeah, the xen is about to die, I think
15:05	<npmr>	bleh
15:05	<warewolf>	npmr- there are some OSes that don't want 512 byte sectors on HDs
15:05	<@caker>	marc_in_lux: prepare for a host reboot soon-ish (tonight?)
15:05	<npmr>	warewolf, heh
15:05	<warewolf>	npmr- I think those were SGI boxes, anyway -- they needed them to be slightly more or less, I forget
15:05	<warewolf>	npmr- and if you didn't frob the drive with that utility, it wouldn't work in the box at all.
15:06	<npmr>	weird
15:06	<npmr>	yeah, these are 36GB barracudas attached to an adaptec dpt-i2o raid card
15:06	<npmr>	U160
15:06	<npmr>	hardware raid
15:06	<warewolf>	the above 5 scsi drives I pasted earlier were on a non-raid U160 adaptec card
15:07	<npmr>	i'm not sure i could get direct access to an individual drive in linux if i attached the drive to that card anyway
15:07	<warewolf>	you could, but it'd be garbledegook.
15:07	<warewolf>	scsi0 : Adaptec AIC7XXX EISA/VLB/PCI SCSI HBA DRIVER, Rev 6.2.8 <Adaptec 29160 Ultra160 SCSI adapter>
15:07	<warewolf>	that's the card
15:07	<npmr>	to tweak the reserved blocks?
15:07	<warewolf>	npmr: oh that's talking to the firmware on the drive. that'd be ok.
15:08	<warewolf>	npmr: what I was saying is the _data_ on the drive would be garbled if you took the drive out of an existing hardware raid, and looked at it under linux.
15:08	<npmr>	honestly, if i had to build that raid all over again, i'd use md raid and a 29160
15:08	<warewolf>	again, I'm not aware of a linux util to muck with the paramaters of the drive.
15:09	<warewolf>	md0 : active raid5 sde1[4] sdd1[3] sdc1[2] sdb1[1] sda1[0]
15:09	<npmr>	that i2o card is a pain in the ass
15:09	<warewolf>	142238976 blocks level 5, 4k chunk, algorithm 2 [5/5] [UUUUU]
15:09	<npmr>	5x 36GB drives?
15:10	<warewolf>	yes
15:10	<npmr>	barracudas?
15:10	\|-\|	linville [~linville@nat-pool-rdu.redhat.com] has quit [Quit: Leaving]
15:11	<warewolf>	4x Vendor: SEAGATE Model: ST336704LC Rev: 6A10
15:11	<warewolf>	1 Vendor: QUANTUM Model: ATLAS V 36 SCA Rev: 0201
15:11	<marc_in_lux>	caker , thanks for announcing it
15:11	<npmr>	ah
15:11	<warewolf>	the atlas is U320 actually
15:11	<marc_in_lux>	how about very soon, or in about 10 hours
15:11	<npmr>	i forget, are the LC drives the full height or half height ones?
15:12	<warewolf>	LP
15:12	<warewolf>	1" tall
15:12	<warewolf>	FH/HH is _huge_
15:12	<@caker>	marc_in_lux: well, if it keeps running I'll leave it be, but it's spitting out "bad PTE" kernel OOPs
15:12	<warewolf>	you realise FH is like almost 4" tall right?
15:12	<warewolf>	and HH is 1.6" tall
15:12	<npmr>	oh, right
15:12	<warewolf>	npmr: your age is showing :P
15:13	<npmr>	ok, so are they HH or 1/2"
15:13	<npmr>	?
15:13	<npmr>	(the LC drives)
15:13	<warewolf>	those drives are LP, low profile, 1" tall.
15:13	<npmr>	ok
15:13	<warewolf>	HH is Half Height, which is 1.6" tall.
15:13	<npmr>	which is what i have
15:13	<warewolf>	FH is full height which is 3.2" tall.
15:13	<marc_in_lux>	just out of interest, do you expect improvement wrt xen somtime soon, caker?
15:13	<npmr>	right
15:14	<@caker>	marc_in_lux: improvements in Xen itself? I hope to pull in the latest Xen Real Soon Now<tm>, since this keeps happening
15:14	<@caker>	marc_in_lux: but it's a little scary -- what if it's worse? :)
15:14	<marc_in_lux>	:-) I was thinking about maybe moving back to UML...
15:15	<warewolf>	npmr: I have a 5-wide SCA chassis for those drives, it's sweet.
15:15	<marc_in_lux>	glad to provide some load there, to beta test
15:15	<@caker>	marc_in_lux: I'll probably just reboot host56 and then get the newest Xen compiled and queued up for the next reboot
15:15	<warewolf>	npmr: fits in the case like it is 3 cdrom drives tall
15:15	<@caker>	marc_in_lux: we'd be happy to move you back, just shoot us a ticket
15:15	<marc_in_lux>	but it seems not really progressing...
15:15	<encode>	ahh, full height drives - i have some 2GB FH scsi drives in a box at home
15:15	<encode>	massive things
15:15	<encode>	nearly the size of my shoe
15:15	<encode>	(i have large feet)
15:15	<npmr>	warewolf, i've got something like that too, except it's a 3-bay gizmo on account of the 1.6" drives
15:16	<warewolf>	npmr: yeah. how mang gig are your HH drives?
15:16	<npmr>	36GB
15:16	\|-\|	dave_ [~chatzilla@64.157.152.10] has quit [Quit: Chatzilla 0.9.68.5 [Firefox 1.0.7/20050915]]
15:16	<npmr>	3-way raid5
15:16	<marc_in_lux>	I'll think about it. The fact that I now have a Linod 256 thanks to the latest upgrade may make UML less sluggish...
15:16	<warewolf>	ouch
15:16	<warewolf>	I hope you have spares :/
15:16	<npmr>	i have two spares
15:16	<npmr>	but that machine isn't even running
15:16	<warewolf>	oh
15:16	<warewolf>	haha
15:17	<npmr>	ha ha indeed :)
15:17	<npmr>	let's see, the array's about 5 years old and the machine is about 10 years old going back through multiple overhauls
15:17	<marc_in_lux>	btw, caker, would you check my billing sometime? I think I should've been billed, but didn't see charges on my CC statement.
15:18	<npmr>	i shut it down maybe six months ago
15:19	<@caker>	marc_in_lux: you paid annually in April of 06, so you're good for another few months
15:19	<marc_in_lux>	yeah, but I had a change of plan in early january
15:19	<marc_in_lux>	and understood I'd be refunded the remainder and billed the new plan
15:19	<marc_in_lux>	I did see the remainder. but not the new plan, yet
15:20	<@caker>	marc_in_lux: you're right -- looks like it should go through tomorrow morning
15:20	<marc_in_lux>	ok, thanks.
15:21	<marc_in_lux>	if i put in a ticket, would you move me to one of the shiny new boxes with the racing stripes that are faster than light?
15:21	<marc_in_lux>	do io_tokens still exist?
15:21	<@caker>	marc_in_lux: sure! .. yes
15:23	<encode>	marc_in_lux: ahaha
15:23	<marc_in_lux>	grmbl - that was the one thing I remembered that really bugged me. but maybe the racing stripes do help there, too.
15:23	<Battousai>	he wouldn't buy em if they didn't
15:23	<@caker>	marc_in_lux: the limiter shouldn't affect anybody during "normal" operation, if it does, let me know
15:24	<marc_in_lux>	back then it was a Linod 80 or 100 or so that I had.
15:25	<marc_in_lux>	I think it mattered when creating a large tar file
15:25	<marc_in_lux>	ticket's in.
15:25	<marc_in_lux>	will I have a 3 digit host number? :-)
15:25	<@caker>	haha, no not yet
15:26	<@caker>	probably in the 60-70 range
15:26	<marc_in_lux>	how many have you got by now?
15:26	<@caker>	76, minus a few
15:26	<marc_in_lux>	great to see your business growing.
15:26	<@caker>	Thanks, yeah -- we should have an east-cast datacenter online in a few weeks, too
15:26	<Darxus>	nice
15:26	<marc_in_lux>	early customers will have preferential share at the ipo, right?
15:27	<Darxus>	caker: where?
15:27	<@caker>	marc_in_lux: haha :)
15:27	<npmr>	so is that pretty much all tasaro does is bring new hosts online and get new customers on them?
15:27	<@caker>	Darxus: atlanta, GA
15:27	<marc_in_lux>	doesn't matter so much to me. Anyway it's behind the horizon to me
15:27	<@caker>	npmr: for now, yes... Massive buildup
15:27	<npmr>	yeah, i see we've gone from 71 to 76 hosts just this afternoon
15:27	<marc_in_lux>	whether it's 6000 km away or 10000 ... doesn't matter a lot.
15:28	<@caker>	that's it for new hosts at HE for a while.
15:29	<@caker>	I think we're going to commit to another cab at TP; mostly to ease the upcoming upgrades
15:29	<Darxus>	how do you handle setting up machines where you don't have any staff?
15:29	<@caker>	and then all new at the atlanta dc
15:29	<npmr>	is there a solution to the lack of power in the TP cabs?
15:29	<Darxus>	have the machines shipped to you, configure them (including IP assigned bo the remote colo) then ship them to the colo?
15:29	<@caker>	TP actually can provide 2x 20A
15:29	<@caker>	HE is 15A only :/
15:29	<npmr>	1x?
15:29	<@caker>	Yup.
15:30	<@caker>	Which gets us a whole 5 machines per cab. yoohoo!
15:30	[~]	npmr makes rude gestures toward fremont
15:30	<@caker>	Darxus: we purchase the components, build, test, ship to DC with idiot-proof labels on everything
15:30	<@caker>	For instance:
15:31	<@caker>	Thanks, I can reach all 5 of them directly now. I am having problems reaching the console on host73, though. Can you verify that host73's serial port is connected to opengear11-18 port 22? Maybe a bad ethernet cable?
15:31	<@caker>	Thanks,
15:31	<@caker>	Thomas Asaro
15:31	<@caker>	Response:
15:31	<@caker>	the power cord was louse, it tighten now.
15:31	<@caker>	thanks.
15:31	<@caker>	Hurricane Electric Support.
15:31	<npmr>	ha ha
15:31	<Darxus>	heh
15:32	<afv-13>	seems they hire only the best
15:36	<npmr>	is there a timeline yet for the replacement of hosts 1-4?
15:36	<@caker>	No, but they will be replaced.
15:36	<marc_in_lux>	caker, I didn't really understand what you did add to this ticket
15:36	<npmr>	this year? this quarter?
15:36	<@caker>	marc_in_lux: that was a note for linode support, ignore it, please
15:37	<@caker>	npmr: I think it's safe to say 1/2 of this year
15:37	<marc_in_lux>	ok - I won't have to change anything, right? Just wait for the prompt to come...
15:37	<@caker>	*first 1/2
15:37	<npmr>	k
15:37	<@caker>	marc_in_lux: yes -- you'll get an email once the migration is set up
15:39	<marc_in_lux>	thanks
15:40	<Darxus>	caker: do you ever have to give colo staff root passwords?
15:40	<@caker>	Darxus: hell no
15:41	<Darxus>	heh, good
15:41	<@caker>	they had access to a few switches at some point, but lost that privilege when they turned off a switchport rather than just null routing an IP
15:41	<npmr>	me too, but they keep asking
15:41	<@caker>	(this was over some abuse issues on a particular IP)
15:41	<Darxus>	do you do static IPing? or is that handled by DHCP similar to the VPSes?
15:41	<@caker>	on my hosts? They're configured statically
15:42	<Darxus>	I'm kind of surprised that's always worked out
15:42	<@caker>	what has?
15:42	<Darxus>	static IP configuration
15:42	<@caker>	for linodes?
15:42	<Darxus>	before shipping, without having to ever have anybody touch it at the colo
15:42	<Darxus>	for the hosts
15:43	<@caker>	Oh, well, I know what ranges we have and keep very, very good configuration documentation
15:43	<@caker>	I know a host's IP, port on which switch, RPC port and console port
16:03	\|-\|	marc_in_lux [~53d99551@webuser.linode.com] has quit [Quit: CGI:IRC (Ping timeout)]
16:04	<iggy>	yeesh, 15A isn't much to work with, even if you are trying to optimize for power consumption that's not going to fill a rack
16:05	<npmr>	2U nuclear reactor
16:05	\|-\|	adamg [~misthos@zeus.misthos.com] has quit [Ping timeout: 480 seconds]
16:05	<iggy>	linode does employ a former nuclear operator
16:09	<@mikegrb>	mmm power
16:10	<iggy>	go whip a 2U nuclear reactor up real quick
16:11	<@mikegrb>	k
16:11	<npmr>	:)
16:12	<@mikegrb>	for tp we don't have to worry about shielding
16:12	<@mikegrb>	or not so much, you know, like russia
16:12	<npmr>	on account of it being texas?
16:12	<iggy>	you're never going to be there, I wouldn't worry about it period
16:12	<Kurt>	all my exes live in Texas...
16:12	<iggy>	mine too
16:14	<npmr>	i can't make any such generalizations
16:14	<npmr>	oh, wait
16:14	<npmr>	i can
16:14	<npmr>	all my exes live in the upper midwest
16:14	<encode>	funnily enough, all my exes live in .au
16:15	<npmr>	my brother's exes all live.... uh.... on earth?
16:15	<Kurt>	Australia is a collectivist piece of shit
16:15	<encode>	huh?
16:16	<Kurt>	just that
16:16	<Kurt>	PEYTON MANNING FOR GOVERNOR!
16:16	[~]	Battousai waits for the joke about exes living in C:
16:16	<Kurt>	encode: you would do well to familiarize yourself with the writings of the eminent 20th-century Russian-American philosopher Ayn Rand
16:17	<encode>	see now i'm confused. why would i possible care about a russian-american philospher?
16:17	<encode>	possibly*
16:18	<Kurt>	because she was never wrong
16:24	<ajmitch>	oh dear
16:27	\|-\|	andrew_j_w [~andrew@82-69-30-171.dsl.in-addr.zen.co.uk] has quit [Quit: Konversation terminated!]
16:30	<SpaceHobo>	iggy: 15A isn't too bad if you're running 240V
16:32	<SpaceHobo>	http://www.angryflower.com/atlass.gif <-- Kurt
16:32	<Kurt>	SpaceHobo: I defy you to refute a single word she ever said or wrote
16:33	<SpaceHobo>	Kurt: the word I chose is "Galt"
16:33	[~]	SpaceHobo refutes it
16:33	<SpaceHobo>	Success!
16:35	\|-\|	tjfontaine [~tjfontain@tjfontaine.noc.oftc.net] has joined #linode
16:35	\|-\|	Kurt [1000@evvlinlwt-nas-07-s448.cinergycom.net] has quit [Quit: Those who think pornography demeans women obviously haven't heard of feminism.]
16:35	<tjfontaine>	I thought so.
16:36	<@mikegrb>	mmm cake
16:36	<SpaceHobo>	http://www.angryflower.com/cake.gif <-- caker
16:46	\|-\|	Kurt [1000@evvlinlwt-nas-07-s329.cinergycom.net] has joined #linode
16:46	<warewolf>	purl, I have really got to keep my resume up to date.
16:47	<warewolf>	whoops wrong network
16:47	<encode>	haha
16:47	<encode>	i was wondering what you were talking about
16:47	<warewolf>	purl is a bot, too.
16:48	<Darxus>	I've seen purl somewhere, long ago
16:52	<SpaceHobo>	purl wa sthe original infobot
16:52	<SpaceHobo>	kevin lenzo set it up in efnet #perl in 96 or 97
16:53	<warewolf>	purl is now an amalgum of a raelly old infobot, and some of my code. :)
16:53	<Darxus>	don't thikn #perl is where I saw it
16:54	<Darxus>	I believe it's been in a bunch of channels
16:54	<warewolf>	Darxus: #perl is it's official home, but it's on the perl network in a couple of channels.
16:54	<Darxus>	perl network?
16:54	<warewolf>	irc.perl.org
16:54	<warewolf>	also known as MAGnet.
17:00	<linbot>	New news from forums: Host22 Upgrade / Migration Instructions in System and Network Status <http://www.linode.com/forums/viewtopic.php?t=2595>; \|\| Host13 Upgrade / Migration Instructions in System and Network Status <http://www.linode.com/forums/viewtopic.php?t=2594>;
17:04	\|-\|	kioma [~kioma@adsl-68-124-184-253.dsl.pltn13.pacbell.net] has joined #linode
17:04	<kioma>	host56 problems?
17:04	\|-\|	taupehat [me@taupehat.com] has quit [Read error: Connection reset by peer]
17:05	\|-\|	marc_in_lux [~53d99551@webuser.linode.com] has joined #linode
17:06	<linbot>	New news from forums: Reboot: host56 (2007-01-31) in Xen Public Beta <http://www.linode.com/forums/viewtopic.php?t=2596>;
17:17	\|-\|	kioma [~kioma@adsl-68-124-184-253.dsl.pltn13.pacbell.net] has quit [Quit: kioma]
17:30	\|-\|	GN [~Miranda@pat.foulston.com] has quit [Quit: buh bye]
17:31	\|-\|	marc_in_lux [~53d99551@webuser.linode.com] has quit [Quit: CGI:IRC (Ping timeout)]
17:34	\|-\|	SpaceHobo [~spacehobo@host-84-9-50-117.bulldogdsl.com] has quit [Quit: restarting ]
17:34	\|-\|	SpaceHobo [~spacehobo@host-84-9-50-117.bulldogdsl.com] has joined #linode
17:35	\|-\|	SpaceHobo [~spacehobo@host-84-9-50-117.bulldogdsl.com] has quit []
17:35	\|-\|	SpaceHobo [~spacehobo@host-84-9-50-117.bulldogdsl.com] has joined #linode
17:53	\|-\|	darkbeholder [darkbehold@c220-239-20-56.belrs4.nsw.optusnet.com.au] has quit [Ping timeout: 480 seconds]
18:06	\|-\|	peter2112 [~0cd236c1@webuser.linode.com] has joined #linode
18:07	\|-\|	Peter21124 [~Peter2112@12-210-54-193.client.mchsi.com] has joined #linode
18:11	\|-\|	peter2112 [~0cd236c1@webuser.linode.com] has quit [Quit: CGI:IRC (Ping timeout)]
18:44	\|-\|	Peter21124 [~Peter2112@12-210-54-193.client.mchsi.com] has left #linode []
18:48	\|-\|	superbeef [~lane@doug.clipmediaservices.com] has joined #linode
18:49	<superbeef>	Yo dudes, my deb linode was on host58 the one that ate it the other day
18:49	<superbeef>	my mysql is all freaked out, i've tried purging mysql-server and mysql-client, flushing apt-'s cache and reinstalling
18:50	<superbeef>	but it still errors out loading libmysqlclienet.so.12
18:50	\|-\|	internat-pc [biteme@c210-49-250-210.ipswc1.qld.optusnet.com.au] has joined #linode
18:50	<internat-pc>	so these "major upgrades" that are comming sone are quite intriguing
18:52	<iggy>	aren't they
18:52	<superbeef>	Ooo what are the upgrades?
18:53	<@mikegrb>	mmm cake
18:53	<iggy>	only the mysterious man of cake knows
18:53	<superbeef>	mikegrb are there snapshot backups of my linode on host58 from a few days ago?
18:54	<superbeef>	or better like 2 weeks ago
18:54	<iggy>	backups?
18:54	<Darxus>	heh
18:54	<internat-pc>	linode dont make backups
18:54	<internat-pc>	like at ALL
18:54	<superbeef>	word
18:54	<internat-pc>	its up to u to do ur backups..
18:54	<superbeef>	well shit happens
18:55	<Darxus>	well I would hope that the content of www.linode.com is backed up, or at least redundantly hosted :)
18:55	<iggy>	superbeef: I don't know if debian does it automatically but have you tried re-running ldconfig, etc.
18:55	<ThE__CroW_>	hmmm... need help setting up multiple ips on my gentoo... read /etc/conf.d/net.example done it the way it says, restarted interface, yet it is only seeing 1 of my 2 ips, and i think also using dhcp
18:56	<internat-pc>	obviously their stuff is backed up.. but the client data isnt.. cause one of the hosts completely died a while ago and everyone lost everything
18:56	<superbeef>	iggy: A+
18:56	<superbeef>	iggy: totalyl worked
18:56	[~]	caker gives iggy a cookie
18:56	<Darxus>	nice
18:57	<internat-pc>	ThE__CroW_ if i remember correctly u have to staticly assign the 2nd ip
18:57	<internat-pc>	dhcp wont give it to u
18:58	<ThE__CroW_>	config_eth0=(
18:58	<ThE__CroW_>	"66.220.1.58 netmask 255.255.255.0"
18:58	<ThE__CroW_>	"66.220.1.59 netmask 255.255.255.0"
18:58	<ThE__CroW_>	)
18:58	<ThE__CroW_>	routes_eth0=(
18:58	<ThE__CroW_>	"default gw 66.220.1.1"
18:58	<ThE__CroW_>	"default gw 66.220.1.1"
18:58	<ThE__CroW_>	)
18:58	<internat-pc>	ah
18:59	<internat-pc>	well if gentoo is anything like debian
18:59	<ThE__CroW_>	i do ifconfig and all i'm getting is 66.220.1.58
18:59	<ThE__CroW_>	i see no eth0:1
18:59	<internat-pc>	it should be config_eth0:0=( first ip details )
18:59	<Darxus>	ifconfig -a
18:59	<internat-pc>	and