TheGrebs.com:IRC Logs:#linode:2003:11:19

---	Log	opened Wed Nov 19 00:00:46 2003
01:07	-!-	shakr [~kenn2@goober.ub3r.org] has joined #linode
02:15	<wap>	Good morning.
02:54	<@guinea-sleep>	sighup: g7 gp
02:54	<sighup>	#G7 stats for gp, the Emperor of New England -- Current Level: 30 \| Time to next level: 0 days, 09:13:34 \| Status: online \| Item Total: 254 \| Total Time Idled: 3 days, 15:47:05
03:16	-!-	guinea-sleep is now known as guinea-work
04:55	-!-	Guest38 is now known as adamgent
04:55	-!-	mode/#linode [+o adamgent] by ChanServ
04:55	-!-	adamgent is now known as adamg
05:00	<@adamg>	sighup: g7 adamg
05:00	<sighup>	#G7 stats for adamgent, the sleepy -- Current Level: 24 \| Time to next level: 0 days, 03:51:15 \| Status: online \| Item Total: 206 \| Total Time Idled: 1 days, 15:36:30
05:36	<@adamg>	sighup: linode avail
05:36	<sighup>	Linode availability -- [Linode 64: 29] [Linode 96: 16] [Linode 128: 2] [Linode 192: 0] [Linode 256: 0]
05:37	<@adamg>	3 L64 gone, I wonder how many will be left by the end of the day
05:43	<inkblot>	moin dawgs
05:44	<inkblot>	how many linodes are there?
05:45	<@adamg>	how do you mean
05:45	<inkblot>	production linode VDSes
05:46	<@adamg>	no idea, there are now 12 hosts
05:46	<inkblot>	hmmm
05:46	<inkblot>	are there very many people using the higher-end packages?
05:47	<inkblot>	i get the impression it's mostly 64s, and some people have moved up a notch or two
05:47	<inkblot>	but not so much beyond that
05:50	<@adamg>	i think there are a few hosts on the higher level
06:30	<shakr>	are linodes getting sold now pretty much by word of mouth?
06:30	<@adamg>	i guess some are and others are through the search engines
06:30	<@adamg>	sighup: linode avail
06:30	<sighup>	Linode availability -- [Linode 64: 29] [Linode 96: 16] [Linode 128: 2] [Linode 192: 0] [Linode 256: 0]
06:31	<@adamg>	sighup: linode forums
06:31	<sighup>	adamg: huh?
06:31	<@adamg>	doh
06:32	<shakr>	i recall finding out through wht but it seems now that demand is exceeding supply
06:33	<@adamg>	that does usually seem to be the case esp. with the L64
07:15	<@mikegrb>	there are ads on k5
07:15	<@mikegrb>	and google text now too
07:16	<@mikegrb>	but don't click a google linode ad!
07:16	<@mikegrb>	you're already a customer and you'll charge caker extra money
07:16	<@mikegrb>	;)
08:13	<@adamg>	k5?
08:15	<@adamg>	sighup: linode forums
08:15	<sighup>	adamg: excuse me?
08:15	<inkblot>	kuro5hin
08:15	<@adamg>	miek sighup needs fixing
08:15	<@adamg>	s/miek/mike
08:17	<@adamg>	sighup: linode avail
08:17	<sighup>	Linode availability -- [Linode 64: 29] [Linode 96: 16] [Linode 128: 2] [Linode 192: 0] [Linode 256: 0]
08:36	<@mikegrb>	yha
08:36	<@mikegrb>	he has forget stuff again
08:36	<@mikegrb>	I really just should make the linode forums request part of the linode plugin rather then relying on the factoid database
08:37	<inkblot>	sighup, stats?
08:37	<sighup>	Since Tue Nov 18 17:57:49 2003, there have been 3 modifications and 23 questions. I have been awake for 14 hours, 39 minutes, 32 seconds this session, and currently reference 12587 factoids. Addressing is in require mode.
08:37	<inkblot>	sweeeeet
08:37	<@adamg>	is it only the linode forum but that disappeares?
08:37	<inkblot>	sighup, 242 sighting is <reply>Bless you, $who!
08:37	<sighup>	OK, inkblot.
08:37	<@mikegrb>	other bits disapear
08:37	<@mikegrb>	randomly
08:38	<@mikegrb>	I really need to change to sql for the storage
08:38	<inkblot>	mikegrb, have you ever heard of Dumont?
08:39	<@mikegrb>	no
08:39	<inkblot>	four year old infobot in #tron on slashnet
08:39	<inkblot>	the whole time with addressing mode off
08:39	<@mikegrb>	ahh
08:39	<@mikegrb>	heh
08:39	<inkblot>	dumont is seriously the worst troll in the channel
08:39	<@mikegrb>	lots of factoids I'm sure
08:39	<@mikegrb>	heh
08:40	<inkblot>	currently reference 127241 factoids.
08:40	<@mikegrb>	thats a lot
08:40	<@mikegrb>	this one here got an import of data
08:41	<@mikegrb>	jargon file, a tech dictionary, and common ports
08:41	<@mikegrb>	sighup: port 25
08:41	<sighup>	mikegrb: what?
08:41	<@mikegrb>	ack
08:41	<inkblot>	yeah, there've been a few of those
08:41	<@mikegrb>	but he has problems, he is a perl infobot clone
08:41	<inkblot>	clone?
08:42	<@mikegrb>	heh I go into #tron and instantly get a thousand ctcp versions
08:42	<inkblot>	so, is lenzo's infobot the original or is it the clone
08:42	<inkblot>	?
08:42	<inkblot>	that's how #tron works
08:42	<@mikegrb>	flooterbuck.sf.net
08:42	<inkblot>	if it's proprietary, you'll be kicked immediately
08:42	<inkblot>	policy
08:43	<@mikegrb>	heh
08:43	<@mikegrb>	sweet
08:46	<@mikegrb>	:(
09:31	<Quik>	/etc/init.d/apf start
09:31	<Quik>	Starting APF:lsmod: QM_MODULES: Function not implemented
09:31	<Quik>	any ideas? :/
09:33	<inkblot>	what's apf?
09:34	<Quik>	a firewall
09:34	<Quik>	makes things simple
09:35	<inkblot>	so i guess it's trying to load kernel modules
09:35	<inkblot>	the linode kernels are not modular
09:35	<Quik>	the thing is, going one version down from it seems to work fine
09:35	<Quik>	and i can't see any specific changes regarding the kernel
09:35	<inkblot>	well then i guess you know what to do
09:36	<Quik>	:<
09:36	<@adamg>	i am still undecided as to whether to use plain iptables or use a firewall program
09:36	<inkblot>	what distribution are you using?
09:36	<inkblot>	i have packaged my own firewall for debian
09:36	<inkblot>	i am using it on every system i run
09:36	<inkblot>	including my linode
09:36	<@adamg>	i use debian
09:37	<inkblot>	# Goose debs
09:37	<Quik>	this is RH9 large, messing about with x stuff :)
09:37	<inkblot>	deb http://debian.movealong.org/goose-debs sarge main contrib
09:37	<inkblot>	ha ha red hat
09:37	<inkblot>	so, anyway
09:37	<inkblot>	apt-get install fiarwahl
09:37	<inkblot>	actually,
09:37	<inkblot>	don't
09:37	<@adamg>	isnt sarge still in dev
09:38	<Quik>	tbh it makes no difference to me, it's 100x less hassle getting an x server to run on that than it is to get one going by upgrading a 70mb debian ditro
09:38	<Quik>	*distro
09:38	<inkblot>	welcome to open source software, adamg
09:38	<inkblot>	a work in progress
09:38	<inkblot>	and yes, the package will install just fine on woody
09:39	<@adamg>	tbh I will probable use plain iptables, as I like to know what is happening, just need time to work out how they work
09:41	<inkblot>	the default configuration for fiarwahl locks out everything except ssh anyway
09:41	<inkblot>	i usually configure first, then install
09:42	<@adamg>	all need open is 22, 25, 80, 443, and the other ports ssh uses 1024 - 512 I think
09:43	<@adamg>	i have sample files for locking down a server pretty much just need to understand them a bit better
09:43	<@adamg>	i also want to allow pings and tracert but have them blocked if they start to hit a DOS type limit
09:44	<inkblot>	oh, was it you who asked about iptables rate limiting in the fora?
09:44	<@adamg>	yes
09:45	<@adamg>	i manage to limit ssh connections, but could not limit icmp properly
09:45	<inkblot>	iptables --blah --blah --blah -m limit --limit 3/second --limit-burst 20 -j ACCEPT
09:45	<inkblot>	iptables -p icmp --limit 3/second --limit-burst 20 -j ACCEPT
09:46	<@adamg>	what does the limit burst do
09:46	<inkblot>	that'll let the first 20 through
09:46	<inkblot>	and then limit them at 3 per second
09:46	<inkblot>	it's a token bucket filter
09:46	<@adamg>	so it will limit it to 3 icmp requests per second
09:46	<inkblot>	once the burst is used up, yes
09:47	<@mikegrb>	hmm
09:47	<@adamg>	what is the default burst?
09:47		* mikegrb adds
09:47	<@mikegrb>	inkblot++
09:47	<inkblot>	5
09:47	<inkblot>	are you familiar with token bucket filters?
09:47	<@mikegrb>	Unknown arg `--limit'
09:47		* mikegrb checks man page
09:47	<inkblot>	er,
09:47	<inkblot>	sorry
09:47	<inkblot>	-m limit
09:48	<@adamg>	so if I was to set the burst to 0 and the limit to 1/hour then only 1 icmp should be allowed though an hour
09:48	<inkblot>	and then it'll recognize --limit
09:48	<inkblot>	yes
09:48	<inkblot>	but that's pretty damn severe
09:48	<@adamg>	just want to test it
09:48	<inkblot>	i see
09:48	<@adamg>	i then need to block all the syn etc
09:48	<inkblot>	i really wish there was a way to have the limit module count bytes instead of packets
09:49	<@adamg>	so what would a sample iptables command look for with limit?
09:49	<inkblot>	<inkblot> iptables --blah --blah --blah -m limit --limit 3/second --limit-burst 20 -j ACCEPT
09:49	<@mikegrb>	oh heh
09:49	<@mikegrb>	-m linit
09:49	<inkblot>	replace --blah with your other criteria
09:49	<inkblot>	i.e. -p icmp
09:50	<inkblot>	-m state --state NEW
09:50	<inkblot>	etc
09:50	<inkblot>	i'm really quite fond of the state module
09:50	<inkblot>	it cooks my emergency bacon
09:51	<@adamg>	i really need to find a good guide to iptables
09:51	<inkblot>	the man page has pretty good coverage of the modules
09:51	<inkblot>	it doesn't really cover the tables and chain traversal as well as it should, though
09:52	<@adamg>	me, now
09:52	<@adamg>	opps
09:55	<@adamg>	it doesnt seem to like this
09:55	<@adamg>	iptables -A -p icmp -m limit --limit 3/second --limit-burst 20 -j ACCEPT
09:55	<@mikegrb>	need INPUT after -A
09:55	<@mikegrb>	to tell it which chain to add it to
09:55	<inkblot>	you need to tell it what chain to add that rule to
09:56	<@adamg>	so what chains are tehre
09:56	<inkblot>	the filter table has three base chains
09:56	<inkblot>	INPUT, OUTPUT, and FORWARD
09:56	<@mikegrb>	INPUT, OUTPUT, FILTER i think
09:56	<@mikegrb>	of fwd
09:56	<@mikegrb>	iptables -L will show you all rules on all chains
09:56	<inkblot>	INPUT gets any packets that are addressed to your system
09:57	<inkblot>	OUTPUT gets any packets that are sent from your system
09:57	<inkblot>	FORWARD gets all the ones that are just passing through
09:57	<@adamg>	so if you wanted to block outgoing ssh you would block 22 and add it to the output chain
09:58	<inkblot>	iptables -A OUTPUT -p tcp --dport 22 -j DROP
09:58	<@adamg>	well that icmp rule did not seem to work
09:58	<@adamg>	I added iptables -A INPUT -p icmp -m limit --limit 1/hour --limit-burst 1 -j ACCEPT
09:58	<@adamg>	and I can still ping it several times
09:59	<inkblot>	add a second rule
09:59	<inkblot>	iptables -A INPUT -p icmp -j DROP
09:59	<@mikegrb>	ahhh
09:59	<@mikegrb>	that makes sense
09:59	<inkblot>	once your firewall is all set up,
09:59	<@mikegrb>	if the limit rules are met it gets accepted
09:59	<inkblot>	you should set the default policy on the INPUT chain to drop
09:59	<@mikegrb>	otherwise it falls through
09:59	<@adamg>	im confused again
09:59	<inkblot>	if you were to do that right now,
09:59	<inkblot>	you'd cut yourself off
10:00	<eurozip>	all linear, processes rules one after the other
10:00	<inkblot>	adamg, rules are checked in the order they appear in the chain
10:00	<@adamg>	ok
10:00	<inkblot>	if none match, then the chain's default policy is applied
10:00	<@mikegrb>	:)
10:00	<inkblot>	iptables -nvL INPUT
10:01	<inkblot>	[inkblot@goose:~][inkblot@goose:~]$ sudo iptables -nvL INPUT
10:01	<inkblot>	Chain INPUT (policy DROP 6 packets, 384 bytes)
10:01	<inkblot>	^-- see here i have the policy set to DROP
10:01	<inkblot>	don't do that yet
10:01	<inkblot>	you'll cut yourself off from your system
10:02	<inkblot>	once you have rules to accept all of your normal traffic,
10:02	<inkblot>	then you can set the policy to DROP
10:03	<@adamg>	ok
10:05	<@adamg>	thats better it allows one ping in then drops the rest
10:05	<@adamg>	now to wait an hour to see if it allows another one in
10:06	<@adamg>	so is it best to specify what you want in and then drop the rest
10:06	<eurozip>	yea
10:06	<eurozip>	you dont have to have a -j DROP for everything
10:06	<inkblot>	adamg, like i said,
10:06	<eurozip>	everything you don't want rather
10:06	<inkblot>	add rules to allow all of your normal traffic
10:06	<inkblot>	and then set the policy to DROP
10:07	<inkblot>	so that anything that's not explicitely ACCEPTed is not allowed
10:07	<@adamg>	so what is needed for SSH then is it just port 22
10:07	<inkblot>	yes
10:07	<inkblot>	iptables -A INPUT -p tcp --dport 22 -m state --state NEW -j ACCEPT
10:07	<inkblot>	^-- accept new connections to port TCP/22
10:07	<@adamg>	what does state do
10:08	<inkblot>	it tracks the state of connections
10:08	<@adamg>	doesnt ssh use a range of ports on udp
10:08	<inkblot>	no
10:08	<inkblot>	i have no idea what you might be refering to
10:08	<inkblot>	but it's certainly not ssh
10:09	<inkblot>	(work) (work)
10:16	<@adamg>	i am going to have to spend a few days getting to understand iptables
10:28	<Quik>	caker's network is up to 40k/sec today ;)
10:28	<Quik>	looks like they installed another ISDN line!
11:16	<@adamg>	sighup: linode forums
11:16	<sighup>	adamg: General Discussion: Shopping Cart Software ; Linux Networking: Portscans & other malicious ... ; Feature Request/Bug Report: Why does my linode keep ... ; General Discussion: what is our connection ... ; General Discussion: Life beyond the forums
11:16	<@adamg>	sighup: linode avail
11:16	<sighup>	Linode availability -- [Linode 64: 27] [Linode 96: 16] [Linode 128: 2] [Linode 192: 0] [Linode 256: 0]
11:50	-!-	jax [~stbe@148.100.214.179] has joined #linode
11:56	<@adamg>	sighup: g7 adamg
11:56	<sighup>	#G7 stats for adamgent, the sleepy -- Current Level: 25 \| Time to next level: 0 days, 04:52:35 \| Status: online \| Item Total: 212 \| Total Time Idled: 1 days, 21:52:30
11:59	<EFudd>	sighup: g7 elfindrol
11:59	<sighup>	#G7 stats for Elfindrol, the Master of teh Idle. -- Current Level: 28 \| Time to next level: 0 days, 07:12:16 \| Status: online \| Item Total: 222 \| Total Time Idled: 2 days, 17:06:50
12:04		* jax is away: out to lunch
12:33	<@adamg>	ok im back to iptables if I have a rule that says it is allowed 1 packet an hour, 3 packets are then sent 2 and 3 get bounced, will after an hour another packet be allowed through or will everything be blocked until the counters are zeroed
12:38	<inkblot>	perhaps you should adjust your parameters so that your test doesn't take so long
12:38	<inkblot>	for example
12:38	<inkblot>	make it 12/minute
12:38	<inkblot>	and you'll know in five seconds whether it works the way you want it to
12:41	<@adamg>	i guess I could do that, but are the counters reset automattically?
12:46	<inkblot>	the counter for that rule is specific to that rule
12:46	<inkblot>	when you remove the rule, the counter no longer exists
12:46	<inkblot>	when you recreate the rule, it's a whole new counter
12:48	<@adamg>	so once the rule for the limit has been reached all packets would then be blocked until the counter is zeroed unsing -Z for the rule is removed
12:48		* jax is back (gone 00:44:15)
12:48	<jax>	sighup: g7 jax
12:48	<sighup>	#G7 stats for jax, the noob -- Current Level: 7 \| Time to next level: 0 days, 00:02:30 \| Status: online \| Item Total: 32 \| Total Time Idled: 0 days, 02:25:00
12:48	<inkblot>	so, an iptables rule basically has two parts
12:49	<inkblot>	1) the criteria
12:49	<inkblot>	2) the action
12:49	<inkblot>	if all of the criteria are met, then the action is taken
12:49	<inkblot>	ACCEPT, DROP, REJECT, and some others are terminating actions
12:49	<inkblot>	i.e. the packet is not processed by any subsequent rules in the chain
12:49	<inkblot>	in your rule,
12:50	<inkblot>	"-m limit --limit 12/minute --limit-burst 1" is part of the criteria
12:50	<inkblot>	if the limit's not exceeded, then that part of the criteria are met
12:51	<inkblot>	provided that the rest are met (-p icmp, for example),
12:51	<inkblot>	then the action (-j ACCEPT) is taken
12:51	<@adamg>	ok
12:51	<inkblot>	if that rule is not triggered, the next rule is processed,
12:51	<inkblot>	which will DROP icmp packets
12:51	<@adamg>	ok
12:52	<Quik>	(17:51:45) (@b0t) Welcome Quik's new player Quik, the Linodian! Next level in 0 days, 00:10:00.
12:52	<Quik>	\o/
12:52	<@adamg>	that bit I understand
12:53	<inkblot>	so, what's the problem?
12:55	<@adamg>	once the limit has been exceded and the packets start to get dropped, is the counter for the rule ever automatically reset, or does it only happen when either the rule is dropped or the counters are zeroed by hand using -Z
12:56	<inkblot>	it's a token bucket filter
12:56	<inkblot>	tokens are added to the bucket at a specific rate
12:56	<inkblot>	for example, 12 per minute
12:57	<inkblot>	the bucket fills until it's full, which is the burst
12:57	<eurozip>	12 are added, and until 60 seconds passes from the first packet, nothing else gets through
12:57	<inkblot>	when a packet comes along,
12:57	<inkblot>	if there is a token in the bucket,
12:58	<inkblot>	the token is removed and the limit module registers that its criteria have been met
12:58	<inkblot>	so,
12:58	<inkblot>	when the bucket's empty,
12:58	<inkblot>	that's when the rule stops matching
12:58	<inkblot>	but tokens are still being added at the specified rate
12:59	<@adamg>	ok, so every minute 12 tokens will be added to the bucket
12:59	<inkblot>	well, until the bucket is full
13:00	<inkblot>	your set your burst to 1, right?
13:00	<@adamg>	yes
13:00	<inkblot>	ok
13:00	<inkblot>	so, your bucket only holds at most 1 token
13:00	<inkblot>	as soon as a single token is added,
13:00	<inkblot>	the adding stops
13:01	<inkblot>	and it's restarted when the number of tokens drops back down below the burst
13:04	<@adamg>	ok
13:05	<inkblot>	so, the simple answer to your question...
13:05	<inkblot>	yes, this all happens automatically
13:05	<inkblot>	i think it's worthwhile to understand the concept, though
13:05	<@adamg>	i understand it now with the bucket thing
13:06	<@adamg>	so the bucket should be equal to or greater than the limit per miuute etc
13:08	<inkblot>	the burst is the number of tokens that will be stored,
13:08	<inkblot>	quite literally the size of the bucket
13:08	<inkblot>	basically, when the bucket is full,
13:08	<inkblot>	that's how many matches will occur before rate limiting starts
13:08	<@adamg>	ok thanks
13:08	<inkblot>	once the bucket is empty,
13:09	<inkblot>	matches will only occur at the rate that new tokens are added
13:10	<@adamg>	well that has got me started at any rate
13:10	<@adamg>	will find out about so of the more instuff later, like logging and specific IP rules etc
13:13	<@caker>	hey guys
13:13	<@adamg>	hi chris
13:13	<inkblot>	moin dawg
13:13	<@caker>	whaddup g
13:14	<inkblot>	corrupting young minds, you?
13:14	<@caker>	can I play too?
13:15	<inkblot>	Yes.
13:16	<inkblot>	for the ultimate in hilarity, set your root account's real name to something stupid.
13:17	<inkblot>	here are some possibilities:
13:17	<inkblot>	Bob Oot
13:17	<inkblot>	Woodrow Debianowski III
13:36	<@caker>	quote from recent LKML post:
13:36	<@caker>	Here, I've rebooted to various elevators and run each for at least a day, and for mm3, I'd have to say that the diffs are tolerable, but the smoothest, most responsive is the deadline version. as still gives an occasional 20 millisecond stutter, and cfq might be 10 milliseconds. Even as is a far cry from the near show stopper 15 to 20 second hangs of the performance in the later 2.4's. Great work guys!
13:36		* caker drools
13:37	<@adamg>	?
13:37	<@caker>	He's trying the new i/o schedulers in 2.7
13:37	<@caker>	er 2.6
13:38	<@adamg>	ok
13:59	<inkblot>	15 to 20 second hangs
13:59	<inkblot>	that's totally what i'm seeing on my machine's RAID
14:00		* inkblot sets 2.4 on fire
14:31	<@adamg>	so whats next on the linode to-do list then chris
14:31	<@caker>	There are a few small bugs and cosmetic things
14:32	<@caker>	by far my biggest priority is eliminating the disk i/o contention
14:32	<@adamg>	it sounds like the o_queue thing worked
14:32	<@caker>	temp fix, really, and it only takes affect after someone reboots
14:32	<@caker>	so lots of potential offenders still around
14:32	<@adamg>	oh
14:33	<@adamg>	but wouldnt any fix require a reboot
14:34	<@caker>	Not if there was some keen host-kernel setting I could tweak, or something like that. But, I'm familar with 2.4's various VM systems to know they don't help
14:34	<@adamg>	it would also be good if the, reboot with linode actually rebooted it
14:34	<@caker>	that's pretty low prio
14:34	<@caker>	at the moment
14:34	<@caker>	but it would be nice
14:35	<@adamg>	so 2.6 then
14:35	<@caker>	or maybe the reboot functionality in Lish first
14:35	<@caker>	for example: "ssh user@hostX.linode.com reboot"
14:35	<@caker>	takes you full swing
14:35	<@adamg>	or both
14:35	<@caker>	eventually
14:36	<@adamg>	so you are still hacking away at 2.6 then
14:37	<@caker>	I'm awaiting moderator approval to the bridge mailing list to get some answers
14:37	<@caker>	otherwise, I don't have a testbed atm -- ordering another box today
14:37	<@adamg>	for
14:37	<@caker>	For me to get 2.6 ironed out
14:38	<@caker>	before I ship it up to HE
14:38	<@adamg>	expensive way to do it
14:38	<@caker>	Not really -- besides, I'd like to be ahead of the curve for once
14:39	<@adamg>	by the time you are ready to send that box to he you will be out of l64
14:40	<inkblot>	sighup, linode avail
14:40	<sighup>	Linode availability -- [Linode 64: 26] [Linode 96: 16] [Linode 128: 2] [Linode 192: 0] [Linode 256: 0]
14:40	<@caker>	Right
14:40	<inkblot>	nice little helper monkey
14:41	<@adamg>	those 26 will be gone within a week
14:41	<@guinea-work>	g7 gp
14:41	<@guinea-work>	er
14:41	<@guinea-work>	sighup: g7 gp
14:41	<sighup>	#G7 stats for gp, the Emperor of New England -- Current Level: 31 \| Time to next level: 0 days, 13:53:34 \| Status: online \| Item Total: 254 \| Total Time Idled: 4 days, 02:33:15
14:43	-!-	guinea-work is now known as guinea-pig
14:45	<@caker>	Will 2.6.0-test* move to 2.6.0, or 2.6.1?
14:46	<@guinea-pig>	it should move to 2.6.0-rc1
14:46	<@guinea-pig>	or soemthing :P
14:46	-!-	eurozip [~zip@66.136.35.17] has quit [Quit: Leaving]
14:46	<@guinea-pig>	i think it's just being called 2.6.0-whatever to get people off their backs about "when will we have a 2.6?"
14:47	<@caker>	There's a really loooong thread about that in lkml
14:47	<@caker>	test, pre, rc, etc
14:47	<@guinea-pig>	debian's calling it 2.5.999-testX
14:48	<@mikegrb>	heh
14:48	<@guinea-pig>	the latest glibc that migrated into testing a few days ago
14:48	<@guinea-pig>	libc6-dev depends on linux-kernel-headers, which is 2.5.999-test7-bk-9
14:50	-!-	eurozip [~zip@adsl-66-136-35-17.dsl.snantx.swbell.net] has joined #linode
14:51	<inkblot>	first it will be -testN
14:51	<inkblot>	then it will be -preN
14:51	<inkblot>	then there will be a 2.6.0
14:51	<@caker>	no rc?
14:51	<inkblot>	nope
14:51	<@caker>	is rc gone?
14:51	<inkblot>	followed by 2.6.1-preN
14:51	<@caker>	ok
14:51	<inkblot>	culminating in 2.6.1
14:51	<@guinea-pig>	heh
14:52	<heidi>	sighup: g7 heidi
14:52	<sighup>	#G7 stats for heidi, the princess of orion -- Current Level: 31 \| Time to next level: 0 days, 06:00:03 \| Status: online \| Item Total: 234 \| Total Time Idled: 4 days, 08:22:35
14:52	<inkblot>	just like every other kernel release
14:52	<heidi>	I am number 1
14:52	<heidi>	heh
14:52	<heidi>	yay
14:52	<@guinea-pig>	erm
14:52	<@guinea-pig>	i thought every other release was -pre then -rc
14:53	<inkblot>	i've never seen a -rc kernel
14:53	<@caker>	2.4.23-rc3
14:53	<tjfontaine>	The latest prepatch for the stable Linux kernel tree is: 2.4.23-rc2
14:53	<@caker>	marcello released 2.4.23-rc3 today
14:54	<inkblot>	ha ha marcello
14:54	<tjfontaine>	website still stays rc2
14:54	<inkblot>	that guy's prone to do just about anything
14:55	<@caker>	heh
14:55	<@caker>	That reaction comes up a lot when his name is mentioned
14:55	<inkblot>	he's a joke
14:55	<@guinea-pig>	sigh
14:55	<inkblot>	i have known his countrymen to hang their heads in shame
14:56	<@guinea-pig>	my boss: do you know what ISO is?
14:56	<@guinea-pig>	merck: international standards organisation
14:56	<@caker>	inkblot: ouch
14:56	<tjfontaine>	International Standards Organizations
14:56	<tjfontaine>	?
14:56	<@guinea-pig>	my boss: no, it's our environmental policy
14:56	<tjfontaine>	heh
14:56	<@guinea-pig>	s/merck/me/
14:56	<@guinea-pig>	me: groan
14:57	<@guinea-pig>	he was referring to ISO14001, but he didn't say that
14:57		* tjfontaine nods
14:57	<eurozip>	idiot
14:57	<eurozip>	heh
14:57	<@guinea-pig>	at least i didn't say "it's a cdrom"
14:57	<tjfontaine>	you should have said, cd image file?
14:57	<tjfontaine>	hehe
15:41	-!-	sighup [~sighup@thegrebs.com] has quit [Remote host closed the connection]
15:49	-!-	sighup [~sighup@webuser.oftc.net] has joined #linode
15:53	-!-	mode/#linode [+o sighup] by adamg
15:53	<@adamg>	rr
15:53	<@adamg>	sighup: rr
15:53	<@sighup>	adamg spins the chamber, pulls the trigger, and lives to hand you the gun.
15:55	<heidi>	rr
15:55	<heidi>	sighup: rr
15:55	<@sighup>	heidi spins the chamber, pulls the trigger, and lives to hand you the gun.
15:55	<@adamg>	sighup: rr
15:55	<@sighup>	adamg spins the chamber, pulls the trigger, and lives to hand you the gun.
16:04		* EFudd gets frustrated with losing his network connection to linode again
16:10	<@caker>	EFudd: still there?
16:27	<eurozip>	tjfontaine, heh
16:29	<EFudd>	NEW YORK - People for the Ethical Treatment of Animals has delayed a
16:29	<EFudd>	new ad campaign with the slogan "Get Neutered, It Didn't Hurt Clay
16:29	<EFudd>	Aiken," while it waits to see if Aiken will apologize for negative
16:29	<EFudd>	comments he made about cats
16:30	<EFudd>	http://smlnk.com/xrcQ
17:00	<Artifex>	caker: you around?
17:00	<@caker>	yes
17:01	<Artifex>	my invoice is weird. :-)
17:01	<@caker>	From last month?
17:01	-!-	jax [~stbe@148.100.214.179] has quit [Quit: home]
17:02	<Artifex>	oh, hrmm, nevermind. :-)
17:02	<@caker>	Yeah, you're carried through until the end of Nov.
17:02	<Artifex>	you prorate EVERYTHING, huh?
17:02	<Artifex>	Additional 3 IP Addresses 10/27/2003 10/31/2003 $0.00 $0.49
17:02	<Artifex>	that was throwing me for a loop. ;-)
17:02	<@caker>	heh
17:03		* Artifex <-- problem customer ;-)
17:03	<Artifex>	:-P
17:03		* Artifex goes back to work
17:03	<@caker>	no such thing as a problem customer :)
17:06	<Quik>	pff
17:06	<Quik>	i tried, but obviously failed then :(
17:06	<@caker>	as long as they pay :)
17:06		* Quik scribbles on list
17:20	<EFudd>	http://www.ebaumsworld.com/mariosplat.html
18:18	-!-	slacker [~You_Wish@68.209.131.3] has joined #linode
18:19	<@adamg>	be back in an hour or so
18:19	<slacker>	anybody awake
18:19	<@adamg>	yes
18:19	<inkblot>	24/7
18:20	<slacker>	?
18:20	<slacker>	in my linode
18:20	<slacker>	who would i check what space is avaiable
18:20	<@caker>	df -h
18:20	<slacker>	how would i check what space is avaiable
18:20	<slacker>	k
18:20	<@adamg>	whats the -h for?
18:20	<@caker>	human readable format
18:20	<inkblot>	human readable
18:20	<@mikegrb>	huuman
18:20	<@mikegrb>	heh
18:20	<@caker>	more human
18:20	<inkblot>	hyoooooman
18:20	<@adamg>	arrah
18:21	<@adamg>	just converts the figures to meg then
18:21	<slacker>	can i paste result
18:21	<@caker>	sure...
18:22	<slacker>	none 92M 0 91M 0% /dev/shm
18:22	<slacker>	ops
18:22	<slacker>	none 92M 0 91M 0% /dev/shm
18:22	<slacker>	hmm
18:22	<slacker>	hmm
18:22	<slacker>	not pasting
18:22	<slacker>	wait
18:22	<inkblot>	it's the /
18:22	<@caker>	hehe
18:23	<@mikegrb>	hah
18:23	<slacker>	./dev/ubda 938M 600M 299M 67% /
18:23	<slacker>	lol
18:23	<slacker>	thaks
18:24	<slacker>	i know i got more than that available
18:24	<slacker>	i have 5gig i think
18:24	<slacker>	total
18:24	<@adamg>	what disk images did you make?
18:24	<slacker>	and only 1gig used
18:24	<slacker>	rh8
18:24	<slacker>	small
18:24	<inkblot>	make more block devices
18:25	<inkblot>	format them
18:25	<inkblot>	mount them
18:25	<@caker>	or, resize the one you have to a larger size
18:25	<inkblot>	yeah, that's probably what i'd do
18:25	<slacker>	i want have like afew webpages
18:25	<slacker>	afew forums
18:26	<slacker>	afew ircd and ip
18:26	<@caker>	sounds good
18:26	<slacker>	i dont know how to break it up so it fits the right way
18:26	<EFudd>	[Wed Nov 19 18:24:52 2003][Wed Nov 19 18:24:52 2003] [notice] child pid 4582 exit signal Segmentation fault (11)
18:26		* EFudd ponders why he's not getting a core
18:26	<EFudd>	ulimit -c is 5000
18:26	<slacker>	i am trying to avoid down tim
18:27	<@caker>	slacker: no way to resize or adding more disks without rebooting
18:27	<slacker>	k
18:27	<@caker>	it only takes a minute or two
18:27	<slacker>	so i want to do everthing at the sametime for one boot if possiable
18:27	<@adamg>	shutdown the linode, resize the disk and then boot
18:27	<@caker>	Shutdown, click on your disk image, edit the size, press save, then boot back up
18:27	<@caker>	:)
18:28	<@caker>	adamg++
18:28	<@caker>	inkblot++
18:28	<slacker>	before i upgrade i did this
18:28	<slacker>	i had a swap 128 i think
18:28	<slacker>	and everything was max out
18:28	<@adamg>	back in 30 mins or so
18:28	<slacker>	and it said it was full before i had done anything
18:28	<@caker>	ok
18:28	<inkblot>	slacker, how much memory do you have?
18:29	<slacker>	192
18:29	<inkblot>	ok
18:29	<inkblot>	i'd increase swap
18:29	<inkblot>	you should definitely have at least as much swap as memory
18:29	<inkblot>	i generally prefer even more
18:29	<inkblot>	double
18:29	<slacker>	so this is over my head
18:29	<slacker>	sorry this is over my head
18:29	<slacker>	let me thing about this more
18:30	<@caker>	slacker: just resize the swap image to between 200-300
18:30	<slacker>	i have clue what you guys are saying
18:30	<slacker>	k
18:30	<slacker>	swap is what?
18:30	<slacker>	memory or ram use
18:30	<@caker>	it's virtual ram
18:30	<@caker>	on the disk
18:30	<slacker>	k
18:31	<@caker>	so it's slow as hell
18:31	<slacker>	k
18:31	<inkblot>	memory and ram are the same thing
18:31	<@caker>	but, gives you more room to run more
18:31	<slacker>	k
18:31	<inkblot>	swap is where your system moves things when space is tight in memory
18:31	<slacker>	k
18:31	<slacker>	let me see somthing
18:31	<inkblot>	and the combination of swap and memory is virtual memory
18:33	<slacker>	128MB Swap Image that is what is on now
18:33	<inkblot>	and you have 192MB of ram?
18:33	<slacker>	ya
18:33	<inkblot>	ok
18:33	<inkblot>	increase the size of your swap image
18:34	<inkblot>	it should be at least the same size,
18:34	<inkblot>	preferrably larger
18:34	<slacker>	so the highest would be what
18:34	<inkblot>	i generally make my swap space double the size of ram
18:34	<slacker>	k
18:34	<slacker>	that is why it didnt work the first time
18:35	<inkblot>	why what didn't work?
18:35	-!-	risto [~risto@64.5.53.140] has joined #linode
18:35	<slacker>	i put 128 and when it was only 128 and it said i was out of memory
18:35	<@caker>	hello risto
18:35	<risto>	hello caker
18:35	<slacker>	let me think more what i need ask
18:36	<slacker>	crap
18:36	<slacker>	thanks
18:36	<slacker>	dont know how to ask the right question dah
18:36	<slacker>	later
18:36	-!-	slacker [~You_Wish@68.209.131.3] has quit [Quit: ��M�� (www.come.to/ircghost) - Using Theme: Plain Theme]
18:36	<inkblot>	damn kids
18:36	-!-	adamg [~misthos@cpc1-cdif1-5-0-cust33.cdif.cable.ntl.com] has quit [Ping timeout: 480 seconds]
18:36	<@mikegrb>	hmm
18:37	<inkblot>	No packages found matching kernel-source-2.6*.
18:37	<inkblot>	:<
18:44		* EFudd pours a load of crap on apache2
18:44		* EFudd steps back to discern said pile
18:44		* EFudd can't tell original from new
18:45	<@guinea-pig>	kernel-source-2.6.0-test9 - Linux kernel source for version 2.6.0-test9 with Debian patches
18:45	<@guinea-pig>	?
18:47	<inkblot>	hmmmm
18:47	<inkblot>	sid?
18:47	<@caker>	you know .. windows doens't handle out of memory very well
18:47	<@guinea-pig>	sarge
18:47	<inkblot>	reeaaaaaalllly
18:47	<@guinea-pig>	windows doesn't handle much very wel
18:48	<EFudd>	forever apache2 # strace -fvpo 4579 apache2.log
18:48	<EFudd>	strace: Invalid process id: o
18:48	<EFudd>	strace: 4579: command not found
18:48		* EFudd smacks that command line parsing
18:48	-!-	slacker [~You_Wish@68.209.131.3] has joined #linode
18:48	<slacker>	i think i got a simpler question my brain can handle if it is ok
18:48	<@caker>	slacker: go for it!
18:49	<slacker>	An error has occurred during installation
18:49	<slacker>	The PHP configuration on your server doesn't support the database type that you chose
18:49		* EFudd runs
18:49	<slacker>	i get that trying to lanch phpbb2
18:49	<EFudd>	what hostos ?
18:49	<inkblot>	what database are you using?
18:49	<slacker>	i think mysql
18:50	<slacker>	ya
18:50	<slacker>	the new one
18:50	<inkblot>	is it mysql or isn't it?
18:50	<slacker>	it is
18:50	<inkblot>	ok
18:50	<inkblot>	have you installed the php-mysql package
18:50	<inkblot>	(or whatever it's called in your distribution)
18:50	<EFudd>	The answer is obvious, you didn't compile php with mysql support. Back to my question though, what host OS.. er.. ok. redhat? /me runs again
18:51	<slacker>	redhat
18:51	<slacker>	8
18:51	<inkblot>	ok, well
18:51		* EFudd doesn't do redhat, Windows, Printers, or Laundry
18:51	<inkblot>	it's been years since i last ran red hat
18:51	<inkblot>	i forget how to look up package names
18:51	<@guinea-pig>	i used redhat for a week once
18:51	-!-	guinea-pig is now known as guinea-lseep
18:51	<slacker>	apt-get i used
18:51	<inkblot>	but there should be a package which installs the php-mysql bindings
18:52	<slacker>	i had a old mysql
18:52	<inkblot>	and you need to find out what that package is called and install it
18:52	<EFudd>	USE="+mysql" emerge mod_php
18:52		* EFudd ducks
18:52	<@caker>	slacker: apt-get install php-mysql
18:52	<slacker>	k
18:53	<@mikegrb>	heh
18:53	<risto>	packaging is overrated, get the source
18:53	<@mikegrb>	portage++
18:53	<@mikegrb>	EFudd: got some cups questions for you heh
18:53	<EFudd>	forever apache2 # egrep -ic 'Segmentation' /var/log/apache2/error_log
18:53	<EFudd>	48
18:53	<@caker>	mikegrb: lol :)
18:53	<@mikegrb>	actually that's a good idea
18:53	<EFudd>	This is gonna be fun....
18:53	<inkblot>	risto, you'll get tired of that pretty quickly
18:53	<@mikegrb>	got a printer not hooked up at all right now real nice
18:53	<@mikegrb>	no printing capabilities at home heh
18:53	<EFudd>	mike, cups has a web interface i think. use it. :-)
18:53	<@mikegrb>	heh
18:54	<@mikegrb>	I know
18:54		* mikegrb goes upstairs to plut printer into home server
18:54	<risto>	inkblot, we'll see, 15 years and still not tired
18:54	<risto>	do it with source and you'll get exactly what you want
18:54	<slacker>	paste ok
18:54	<slacker>	Preparing... ########################################### [100%]
18:54	<slacker>	file /usr/bin/comp_err from install of mysql-3.23.56-1.80 conflicts with file from package MySQL-devel-4.0.16-0
18:54	<slacker>	file /usr/bin/mysql_config from install of mysql-3.23.56-1.80 conflicts with file from package MySQL-devel-4.0.16-0
18:54	<slacker>	Artifex: Sub-process /bin/rpm returned an error code (6)
18:55	<EFudd>	Heh.
18:55	<inkblot>	http://tastytronic.net/main/punkin.png <-- risto's preferred method of making a jack-o-lantern
18:55	<EFudd>	example of why I hate redhat, albeit it a very fixable issue.
18:55	<EFudd>	slacker, remove your mysql-devel4, which you don't need, and install the mysql3 package.
18:55	<slacker>	that is what i had before
18:55		* EFudd shrugs
18:55	<slacker>	and got that same error
18:56	<EFudd>	This is why you shouldn't use redhat.
18:56	<EFudd>	Actually, you /should/ use redhat.
18:56		* EFudd crawls back in hole
18:56	<slacker>	yup
18:56	<EFudd>	Did I mention, please put me on /ignore.
18:56		* mikegrb pets the printer
18:56	<slacker>	i just removed everything and put everthing new
18:57	<@guinea-lseep>	if you don't know how bad it can be, you'll never know how god you could have it.
18:57	<slacker>	and that is were i am at
18:57	<@mikegrb>	nice printer with irda for visor, duplexer, and card slots for digital photos
18:57	<EFudd>	Apache2 doesn't appear to listen to "minservers1, maxservers1"
18:58	-!-	guinea-lseep is now known as guinea-sleep
18:59	<risto>	any opinions on godaddy.com registrar services?
18:59	<risto>	worth swapping from register.com?
18:59	<@mikegrb>	risto: I've used them for a domain
18:59	<@guinea-sleep>