--- Day changed --- Log opened Mon Apr 16 23:59:01 2007 00:03 -!- Dreamr_3 [~Dreamer3@74-129-108-228.dhcp.insightbb.com] has joined #linode 00:11 -!- hojuruku [~lmc@ppp508A.dsl.pacific.net.au] has left #linode [] 00:39 -!- cout [~cout@c-68-58-222-12.hsd1.sc.comcast.net] has quit [Ping timeout: 480 seconds] 00:42 -!- cout [~cout@c-68-58-222-12.hsd1.sc.comcast.net] has joined #linode 00:46 < Smark> root@none:/usr/local/bin # ./php -i | grep php.ini 00:46 < Smark> Configuration File (php.ini) Path => /usr/local/lib 00:46 < Smark> root@none:/usr/local/bin # cd /usr/local/lib 00:46 < Smark> root@none:/usr/local/lib # ls 00:46 < Smark> php python2.4 00:46 < Smark> wtf. 00:46 < Smark> how can php not have a config file 00:47 < encode> which distro? 00:47 < Smark> ubuntu 6.06 00:47 < encode> # ls /etc/php5/apache2/ 00:47 < encode> php.ini 00:48 < encode> of course, i dont think php5 is installed by default 00:48 < Smark> its not, any way to verify that that is the correct php.ini and not some sample or whatever 00:48 < encode> but i would usually expect php.ini to be in /etc/php... 00:48 < encode> not sure, im no php expert 00:48 < encode> it usually Just Works (TM) 00:49 < encode> is php working? 00:49 < Smark> yes 00:49 < encode> cos theres some special variable that will output all the php settings 00:49 < encode> http://au2.php.net/phpinfo 00:50 < encode> try that example 00:50 < encode> it should tell you which php.ini php is using 00:50 < ajmitch_> so why do you have a /usr/local/bin/php ? 00:50 < Smark> i dont, i have a /usr/local/lib/php 00:50 < ajmitch_> 16:46 < Smark> root@none:/usr/local/bin # ./php -i | grep php.ini 00:51 < ajmitch_> you were running php from /usr/local/bin 00:51 < Smark> i have no idea 00:51 < ajmitch_> the only way would be that you compiled it yourself 00:52 < Smark> i did 00:53 < encode> well then, there probably isnt a php.ini file 00:53 < Smark> then where do you make changes? 00:54 < ajmitch_> the installation documentation in the php source will probably tell you 00:54 < Smark> alright 00:57 < Xel> Smark, bah. If you want to find all php.ini files.... 00:57 < Xel> sudo updatedb 00:57 < Xel> locate php.ini 00:57 < Xel> (I like locate better than find) 00:58 < Smark> got like 9 =) 00:58 < mikegrb> why not just install the ubuntu package? 00:58 < Smark> does it merge with apache and everythign 100% ok? 00:59 < Xel> If you installed the ubuntu apache package 00:59 < Xel> hehe 00:59 < Smark> lol 00:59 < mikegrb> lolz 00:59 < encode> theres a reason ubuntu has a package management system 00:59 < encode> if you're not going to use it, there are probably better distros to use 01:00 < encode> not using ubuntu's packages for part of the LAMP stack means you pretty much need to compile and maintain the whole lamp stack your self 01:00 < encode> rather than relying on other people's hard work and testing 01:02 < Smark> http://spectralcoding.com/phpinfo.php ugh i compiled it using '--with-php-config=/usr/local/lib/php' and the variables still say its in /usr/local/lib 01:08 -!- FireSlash [~FireSlash@mobile-166-217-096-020.mycingular.net] has quit [Quit: Leaving] 01:35 < encode> is there a make config or something like that? 01:36 < ajmitch_> was there some good reason not to use php5 as supplied in dapper? 01:41 -!- msmiffy [~msmiffy@prospero.kbc.net.au] has joined #linode 01:41 < Smark> nevewr i got it working, i guess i just had to coppy php.ini-dist to /usr/local/lib/php/php.ini and it worked 01:49 < kevincody> woot! arla running! 01:49 < ajmitch_> yay 01:51 < linbot> New news from forums: Great Service in Customer Testimonials 01:53 < Smark> jesus that bot was fast 01:53 < Smark> i posted it like 10sec ago 01:57 < linbot> New news from forums: Great Service & Support in Customer Testimonials 02:04 < msmiffy> Off-topic, I know, but since there are all these clever people here... 02:04 < msmiffy> ...what helpdesk ticketing system creates tickets with serial numbers starting KMM? 02:05 < msmiffy> I've had them from both Yahoo and Telstra (Australian national telco). 02:05 * ajmitch_ shrugs 02:06 < ajmitch_> ugh, telstra :) 02:07 < msmiffy> Yeah. I do notice that the two companies using this never come back with sensible answers or read the mails in question - whether the software is to blame... 02:11 * msmiffy notices that the numbers from both companies are using /KMM\d{8}V\d{5}L0M/ 02:26 -!- Phase [~oh@c-68-52-184-14.hsd1.tn.comcast.net] has quit [Read error: Connection reset by peer] 02:30 -!- jon1 [~jon@pool-71-117-246-196.ptldor.fios.verizon.net] has joined #linode 02:31 -!- jon1 [~jon@pool-71-117-246-196.ptldor.fios.verizon.net] has quit [] 02:36 -!- Smark [~Smark@cm-24-121-61-66.payson.az.npgco.com] has quit [] 03:03 -!- kevincody [nobody@c-24-147-111-136.hsd1.ma.comcast.net] has left #linode [Leaving] 03:08 -!- angie [~3a6bf539@webuser.linode.com] has joined #linode 03:13 -!- zunzun [~18c40415@webuser.linode.com] has joined #linode 03:13 -!- angie [~3a6bf539@webuser.linode.com] has quit [Quit: CGI:IRC (Ping timeout)] 03:16 < zunzun> Just a note to say "hi" - up at two AM for no reason and back to bed in a few minutes 03:16 < msmiffy> 2AM? Nearly teatime here. Hi anyway. 03:16 < zunzun> 2AM in Birmingham, Alabama USA 03:16 < msmiffy> 16:46, Kadina, South Australia 03:17 < msmiffy> Then again, it's 00:16 at my primary Linode. 03:18 -!- getsmart [~getsmart@81-174-45-50.f5.ngi.it] has joined #linode 03:19 < zunzun> Grr- Google Maps doesn't have hi-res images of Kadina 03:19 < msmiffy> No, the Pentagon never thought it worth a bomb ;-) 03:22 < msmiffy> Oh, how stupid. The only way I can connect to another server is to close Chatzilla! 03:22 -!- msmiffy [~msmiffy@prospero.kbc.net.au] has quit [Quit: Smiffy has left the building. http://www.smiffysplace.com] 03:22 -!- msmiffy [~msmiffy@prospero.kbc.net.au] has joined #linode 03:24 -!- zunzun [~18c40415@webuser.linode.com] has quit [Quit: CGI:IRC (Ping timeout)] 03:35 -!- andrew_j_w [~andrew@82-69-30-171.dsl.in-addr.zen.co.uk] has joined #linode 03:38 -!- SpaceHobo [~spacehobo@host-84-9-49-210.bulldogdsl.com] has joined #linode 03:43 -!- msmiffy [~msmiffy@prospero.kbc.net.au] has quit [Quit: Smiffy has left the building. http://www.smiffysplace.com] 03:43 -!- mikegrb [~michael@mail.thegrebs.com] has quit [Quit: Changing server] 03:44 -!- mikegrb [~michael@mail.thegrebs.com] has joined #linode 03:44 -!- mode/#linode [+o mikegrb] by ChanServ 04:47 -!- yumyum [~yumyum@user-3e8880cf.telcl22.dsl.pol.co.uk] has joined #linode 04:47 < yumyum> um 04:48 < yumyum> yum 04:48 < yumyum> yumyum 04:48 < yumyum> exit 04:48 -!- yumyum [~yumyum@user-3e8880cf.telcl22.dsl.pol.co.uk] has quit [] 04:48 < ajmitch_> strange people 04:50 < encode> quite 04:59 -!- getsmart [~getsmart@81-174-45-50.f5.ngi.it] has quit [Quit: Ex-Chat] 05:10 -!- skule [~svs@cpe.atm2-0-12843.0x50a66336.arcnxx10.customer.tele.dk] has joined #linode 05:30 -!- SpaceHobo [~spacehobo@host-84-9-49-210.bulldogdsl.com] has quit [Ping timeout: 480 seconds] 05:56 -!- D[a]rkbeholder [darkbehold@c220-239-20-56.belrs4.nsw.optusnet.com.au] has joined #linode 06:03 -!- darkbeholder [~darkbehol@c220-239-20-56.belrs4.nsw.optusnet.com.au] has quit [Ping timeout: 480 seconds] 06:07 -!- besonen_mobile_ [~besonen_m@71-220-227-185.eugn.qwest.net] has joined #linode 06:09 -!- besonen_mobile [~besonen_m@71-220-225-182.eugn.qwest.net] has quit [Ping timeout: 480 seconds] 06:35 -!- SpaceHobo [~spacehobo@yttrium.canonical.com] has joined #linode 06:58 -!- cout [~cout@c-68-58-222-12.hsd1.sc.comcast.net] has quit [Read error: Operation timed out] 07:10 -!- cout [~cout@c-68-58-222-12.hsd1.sc.comcast.net] has joined #linode 07:31 -!- Dreamr_3 [~Dreamer3@74-129-108-228.dhcp.insightbb.com] has quit [Quit: This computer has gone to sleep] 07:52 -!- schultmc_ [~schultmc@c-68-58-138-203.hsd1.in.comcast.net] has joined #linode 08:22 -!- Achindra [~achindra@203.199.144.195] has quit [Ping timeout: 480 seconds] 08:25 -!- linville [~linville@azure.tuxdriver.com] has joined #linode 08:50 -!- tibbetts [~tibbetts@c-24-34-111-180.hsd1.ma.comcast.net] has quit [Quit: tibbetts] 09:09 -!- getsmart [~getsmart@81-174-45-50.f5.ngi.it] has joined #linode 09:10 -!- tibbetts [~tibbetts@onat.streambase.com] has joined #linode 09:18 -!- Pryon [~Pryon@c-76-18-90-150.hsd1.nm.comcast.net] has quit [Quit: Leaving] 09:44 -!- linville [~linville@azure.tuxdriver.com] has quit [Read error: Operation timed out] 09:50 -!- TJF [~MirandaIM@pat.foulston.com] has joined #linode 09:51 < caker> mmmm Compress::Bzip2 09:52 < caker> perldoc Compress::Bzip2 09:52 < npmr> TP sucks 09:53 < caker> it was a little scary reading their Comments and Suggestion forum yesterday 09:54 < caker> but I imagine that forum's primary purpose is for bitchfests in the first place... 09:55 < troy> are TP still stalling ? 09:55 < caker> but they do seem a little stressed under, what I can only imagine, the strain of the merger (who knows) 09:55 < caker> troy: they say cage built by the end of the week, then no ETA for gear installed. 09:56 -!- linville [~linville@azure.tuxdriver.com] has joined #linode 09:56 < caker> Hopefully after my comments to them (nice long, pointed comments) they're telling us it'll take longer than it actually will 09:57 < caker> I don't think any amount of complaining/bitching will speed things up, so I'm just giving up for now. 09:57 < caker> I'll start riding them more towards the end of the week 09:57 < troy> ok my turn now... 09:57 < troy> TP sucks 09:57 < troy> ;) 10:03 < troy> once of my clients has a dedicated box at TP.. took over 2 weeks just to get them to put a debian image on it.. and hounding them every few days did not help 10:05 < tasaro> The false estimates on their part is the worst... because that's what we base our estimates on for announcements 10:06 < tasaro> So they fix that problem by providing *no* eta now... Guess they can't be accused of missing their target that way 10:17 -!- wastrel [~wastrel@76.8.67.2] has joined #linode 10:53 < caker> KILL DASH NINE, NO MORE CPU TIME 10:54 < scott> is this a good time to ask about TP and the new servers? 10:54 < scott> :) 10:57 < caker> superglue is damn sticky stuff 10:57 < caker> I should probably just by a new cell phone... This one's been through hell and back. Including a 60mph road surfing session (fell out of my unzipped pocket on the motorcycle) 10:57 < caker> *buy 10:58 < npmr> i pulled mine out yesterday and someone said it looked like it'd been through a war 10:58 < caker> I'm determined to use it until it completely falls apart, however 10:58 < npmr> likewise 10:58 < npmr> or until i figure out i don't need a cell 11:00 < wastrel> i don't like mine 11:03 < Xel> caker, I'm tired of your datacenters pushing Linode around :\ 11:04 < Xel> Someone needs to lay the royal smackdown on a CEO or two 11:05 < scott> right, lets march up to their corporate buildings carrying torches and pitch forks 11:06 < Xel> Seriously. They abuse Linode quite a bit. 11:06 < caker> we've got a fair amount of bandwidth, cumulatively, to throw their way 11:06 < caker> not that there's anything WRONG with that 11:06 < Xel> caker, emails are easy to ignore/delete. 11:09 < Xel> As are apparently IRC messages 11:09 < Xel> :P 11:09 < caker> who, what? 11:09 < caker> where am I? 11:09 < wastrel> hello 11:09 < Xel> ... you are in the IRC channel for my company. 11:09 < Xel> Now, remind me of the bank number and PIN... 11:10 < Xel> So that I can count my money 11:11 < opello> heh 11:12 < Xel> It didn't work 11:36 -!- schultmc_ [~schultmc@c-68-58-138-203.hsd1.in.comcast.net] has quit [Quit: Client exiting] 12:00 -!- cmantito [~gphreak@c-76-98-50-187.hsd1.nj.comcast.net] has joined #linode 12:05 < kvandivo> i'd like to see a picture of a cell phone that looks like it has been through a war. specifically, npmr's. 12:14 -!- Dreamr_3 [~Dreamer3@74-129-108-228.dhcp.insightbb.com] has joined #linode 12:15 -!- arrummzen [~arrummzen@c-67-181-229-123.hsd1.ca.comcast.net] has quit [Read error: Operation timed out] 12:17 -!- FireSlash [~FireSlash@mobile-166-217-096-020.mycingular.net] has joined #linode 12:38 -!- eric [~eric@MTRLPQ02-1178072105.sdsl.bell.ca] has joined #linode 12:38 -!- eric [~eric@MTRLPQ02-1178072105.sdsl.bell.ca] has left #linode [] 12:40 -!- FireSlash [~FireSlash@mobile-166-217-096-020.mycingular.net] has quit [Read error: Connection reset by peer] 12:53 -!- jdub [~jdub@home.waugh.id.au] has quit [Ping timeout: 480 seconds] 12:55 -!- Dreamr_3 [~Dreamer3@74-129-108-228.dhcp.insightbb.com] has quit [Quit: This computer has gone to sleep] 13:02 * wferrel Grabs his torches and a lighter... What's TP's address again? 13:05 < npmr> 1333 Stemmons Freeway 13:05 < npmr> Suite #110 13:05 < npmr> Dallas, TX 75027 13:05 < SpaceHobo> We're all going to die in a great big chemical fire. 13:05 < wastrel> starting when? 13:06 -!- getsmart [~getsmart@81-174-45-50.f5.ngi.it] has quit [Remote host closed the connection] 13:08 -!- clanehin [~clanehin@69.134.236.134] has joined #linode 13:17 -!- megaversal [~455d7f80@webuser.linode.com] has joined #linode 13:20 -!- sergiobdt [~sergiobdt@2.121-224-89.dsl.completel.net] has joined #linode 13:23 -!- afv-13 [~afv@rkdp-ip-nas-1-p120.telkom-ipnet.co.za] has joined #linode 13:30 -!- adamg [~misthos@zeus.misthos.com] has joined #linode 13:34 -!- sergiobdt [~sergiobdt@2.121-224-89.dsl.completel.net] has quit [Quit: sergiobdt] 13:39 -!- Bdragon [~bdragon@sfDIAL-104.216-254-227.iw.net] has joined #linode 13:58 < SpaceHobo> npmr: hey, do you know why I don't seem to have a postfix verify daemon running? 13:58 < SpaceHobo> npmr: I tried giving it a btree to store its information in but it doesn't create it the way the readme says 14:00 < npmr> that's odd 14:00 < npmr> master should be running that 14:00 < SpaceHobo> [nick@frotz(/var/cache/postfix)] ps fauxw | grep verify 14:00 < SpaceHobo> nick 585 0.0 0.1 2716 732 pts/10 S+ 19:00 0:00 | \_ grep --color=auto verify 14:01 < SpaceHobo> postconf doesn't show me anything fishy 14:01 < npmr> your verify service hasn't mysteriously disappeared from master.cf, has it? 14:01 < SpaceHobo> and it's in the master.cf 14:01 < npmr> hmmm 14:01 < SpaceHobo> [nick@frotz(/etc/postfix)] grep verify master.cf 14:01 < SpaceHobo> verify unix - - - - 1 verify 14:01 < npmr> if you cold start postfix, are there any telling logs? 14:01 < SpaceHobo> nope 14:01 < npmr> i have no idea, man 14:02 -!- adamg_mk2 [~misthos@zeus.misthos.com] has joined #linode 14:02 < npmr> actually, mine's not running either 14:02 < SpaceHobo> Apr 17 19:02:15 frotz postfix/master[1070]: daemon started -- version 2.3.3, configuration /etc/postfix 14:02 < SpaceHobo> odd 14:02 < SpaceHobo> I'm using 2.3.3 14:03 < npmr> ii postfix 2.3.8-2+b1 A high-performance mail transport agent 14:03 < npmr> i went to etch 14:03 < SpaceHobo> I'm edgy 14:03 < npmr> ok 14:03 < SpaceHobo> but yeah, something fishy in 2.3 14:07 -!- adamg [~misthos@zeus.misthos.com] has quit [Ping timeout: 480 seconds] 14:08 < SpaceHobo> npmr: I ran verify manually and got Apr 17 18:59:32 frotz postfix/verify[503]: fatal: service verify requires a process limit of 1 14:08 < SpaceHobo> not sure what that means 14:11 < SpaceHobo> npmr: I'm thinking nothing is being verified with my config 14:11 < SpaceHobo> and that's why it's not starting up 14:15 < SpaceHobo> npmr: postfix 2643 0.0 0.3 4620 1932 ? Ss 19:14 0:00 \_ verify -l -t unix -u -c 14:15 < SpaceHobo> reject_unverified_recipient, 14:15 < SpaceHobo> warn_if_reject reject_unverified_sender, 14:15 < SpaceHobo> (for testing) 14:16 < SpaceHobo> npmr: I've recently ditched all greylisting, because it's pretty much broken by design 14:17 < SpaceHobo> I'm switching instead to aggressive sender verification 14:29 -!- Bdragon28 [~bdragon@sfDIAL-104.216-254-227.iw.net] has joined #linode 14:29 -!- Bdragon [~bdragon@sfDIAL-104.216-254-227.iw.net] has quit [Read error: Connection reset by peer] 14:29 -!- Bdragon28 is now known as Bdragon 14:32 < npmr> sorry, had a phone call 14:33 < npmr> SpaceHobo, greylisting is broken how? 14:33 < SpaceHobo> npmr: the fact that some senders need whitelists for a start, and the fact that there are some delivery/timeout harmonics that can block all mail forever for multi-MX sites 14:34 < npmr> ok 14:34 < SpaceHobo> basically it creates a few funny deadlocks in the SMTP 14:42 -!- adamg_mk2 [~misthos@zeus.misthos.com] has quit [Ping timeout: 480 seconds] 14:49 < npmr> how is the aggressive verification working out? 14:57 < SpaceHobo> well I'm doing warn if reject now 14:57 < SpaceHobo> partly to seed the verification btree 14:58 < SpaceHobo> and the only things to warn about verification rejects are things that ought to have rejected on bogus HELO and so forth 14:58 < SpaceHobo> so I'll have to look into some of that 14:58 < npmr> ebay has broken HELOs 14:58 < SpaceHobo> oh? 14:58 < SpaceHobo> how broken? 14:58 < SpaceHobo> non-fqdn? 14:59 < npmr> Apr 17 14:18:50 dorothy postfix/smtpd[23791]: NOQUEUE: reject: EHLO from mxsmfpool09.ebay.com[66.135.209.206]: 453 4.7.1 : Helo command rejected: Host not found; proto=SMTP helo= 14:59 < npmr> (one of many) 14:59 < SpaceHobo> are you sure that's not just spam? 14:59 < npmr> look at the rdns 14:59 * SpaceHobo shrugs 14:59 < npmr> also, whois on the ip 14:59 < npmr> NetRange: 66.135.192.0 - 66.135.223.255 14:59 < npmr> CIDR: 66.135.192.0/19 14:59 < npmr> NetName: EBAY-1 14:59 < SpaceHobo> huh 14:59 < npmr> broken 15:00 < Bdragon> Technically you aren't supposed to hold stupid HELOs against the sender... 15:00 < SpaceHobo> oh well 15:00 < SpaceHobo> welcome to 2007 15:00 < SpaceHobo> you have to work to get mail delivered these days 15:01 < npmr> some days i wonder if any of those messages from ebay are important 15:01 * Bdragon DOES reject traffic if the remote server tries to say it's me, though ;) 15:01 < npmr> but i inevitably just shrug 15:04 -!- dc0e [~dc0e@c-69-243-126-197.hsd1.md.comcast.net] has quit [Ping timeout: 480 seconds] 15:08 -!- blackasylum [~tatanus@li13-36.members.linode.com] has joined #linode 15:08 < npmr> Message-ID: <-- emad 15:09 < afv-13> what do you guys think of uceprotect? 15:10 < afv-13> i added it to my blacklist list but am a little worried by how strict they are 15:12 < Battousai> what you could do is just use it for scoring with spamassassin or something 15:12 < Battousai> if it gets out of hand, lower the score 15:13 < afv-13> good idea, maybe a really stripped down SA wouldn't be too strenuous 15:17 -!- dc0e [~dc0e@c-69-243-126-197.hsd1.md.comcast.net] has joined #linode 15:27 -!- SpaceHobo [~spacehobo@yttrium.canonical.com] has quit [Ping timeout: 480 seconds] 15:44 -!- blackasylum [~tatanus@li13-36.members.linode.com] has quit [Quit: BitchX: now with 42 percent more random quit messages!] 15:50 < afv-13> are mail servers expected to allow icmp 0? 15:52 < npmr> no 15:53 < npmr> not per spec, anyway 15:53 < afv-13> thanks 16:16 -!- smccoy [~smccoy@peer.sef.marchex.com] has joined #linode 16:16 -!- smccoy is now known as tag 16:16 < tag> so 16:16 < tag> if I've got a an old ubuntu breezy box - I know these linode images are diverge from the default distributions a fair amount, is it safe to just upgrade? 16:17 < tag> from breezy to dapper, then dapper to edgy so I can hit the LTS release? 16:18 < caker> tag: they're pretty much straight up installs, cept for a few minor changes (fstab, inittab to start a getty on tty0, and some device nodes that UML wants) 16:18 < tag> I see 16:18 < afv-13> dapper is the LTS 16:18 < afv-13> not edgy 16:18 < tag> so I should be able to just do my upgrades through apt 16:18 < tag> oh right 16:18 -!- SpaceHobo [~spacehobo@host-84-9-49-210.bulldogdsl.com] has joined #linode 16:19 < tag> are you sure 16:19 < tag> 6.10 is the LTS release, edgy is 6.10 16:19 < tag> oh 16:19 < tag> no, you're right 16:19 < tag> 6.06 is the long term release 16:19 < caker> yup 16:19 < afv-13> tag: so update your /etc/apt/sources.list aptitude dist-upgrade (might have to run it 2ce) and hope for the best 16:19 < tag> fuck that, I want edgy anyway 16:20 < caker> tag: good luck 16:20 * caker inserts Linode support disclaimer here 16:20 < afv-13> just don't try skip a release in the upgrade 16:20 < caker> let us know how it goes 16:20 < SpaceHobo> tag: feisty will release on thursday, and will be an LTS as well 16:20 < afv-13> no it wont 16:21 < SpaceHobo> ?? 16:21 < caker> SpaceHobo: he told you 16:21 < afv-13> it'll be a standard 18 monther 16:21 < tag> yeah feisty is not an LTS release 16:21 < tag> but I already run feisty on my workstations 16:22 < SpaceHobo> oh maybe it's 7.10 that will be LTS 16:22 < SpaceHobo> not 7.04 16:22 < afv-13> i think gibbon+1 might be 16:23 < SpaceHobo> hmmm 16:23 < tag> cute, apt is segfaulting 16:24 < SpaceHobo> I can't find any public mention of which post-dapper release will be LTS 16:24 < SpaceHobo> so I guess nothig is decided yet 16:30 * caker milks TP tech support guy for insider info 16:30 < caker> nothing substantial... basically what I thought. He's been there > 1.5 yr, 4 months since the merger, lots of prodcedural changes, support quality up/down - kinks to work out, etc. 16:30 < caker> blah blah .. no updates with regard to us 16:30 < tasaro> heh.. 16:31 < Battousai> ? 16:31 < npmr> tag, delete the cache files 16:31 < npmr> tag, /var/lib/cache/apt/*.bin 16:31 < SpaceHobo> oh yeah 16:31 < SpaceHobo> pkgcache.bin corruption 16:31 < npmr> sorry 16:31 < caker> sudo rm /var/cache/apt/pkgcache.bin /var/cache/apt/pkgsrccache.bin , methinks 16:31 < SpaceHobo> only ever seem to get that on my linode 16:31 < npmr> wrong path 16:31 < npmr> likewise 16:31 < SpaceHobo> suspicious 16:31 < caker> SpaceHobo: looks like it does exists elsewhere: https://bugs.launchpad.net/ubuntu/+source/apt/+bug/57393 16:33 < caker> maybe not... 16:36 -!- getsmart [~getsmart@81-174-45-50.f5.ngi.it] has joined #linode 16:39 -!- tehdan [~dan@omicronpersi8.plus.com] has quit [Ping timeout: 480 seconds] 17:04 -!- linville [~linville@azure.tuxdriver.com] has quit [Quit: Leaving] 17:04 < linbot> New news from forums: Looking for a decent postifx + spam filtering debian guide. in Email/SMTP Related Forum 17:13 -!- rk [~prims@gateway.abk-aalborg.dk] has joined #linode 17:15 -!- getsmart [~getsmart@81-174-45-50.f5.ngi.it] has quit [Quit: Ex-Chat] 17:21 < rk> Hi. I have a linode at the Dallas hosting center and am considering to migrate my linode to another datacenter to get the hardware upgrade. But I not sure how it would affect my linode. I currently have two ip's and I would need to still need have two (changing them would not be a problem), is this possible? I also read something about blocked ports at the new datacenter, and I found a thread in the forums listing at least two blocked ports that I need (a 17:21 -!- andrew_j_w [~andrew@82-69-30-171.dsl.in-addr.zen.co.uk] has quit [Read error: Connection reset by peer] 17:26 < tierra> rk: I imagine if they have room at HE in Fremont, they could migrate you there 17:26 < iggy> rk: the dallas DC is getting updates soon too 17:26 < iggy> *upgrades 17:26 < tierra> though TP upgrades should be available in the next week or so 17:27 < iggy> how soon depends on a number things (mostly tp) 17:27 < tierra> yes 17:27 < rk> so atlanta and HE is the only ones that has received the upgrades? 17:27 < tierra> yes 17:28 < tierra> leaving just Dallas 17:29 < rk> Maybe I should just be patient then. The thread at http://www.linode.com/forums/viewtopic.php?t=2634&start=30 does not suggest a positive outlook however. 17:29 < tierra> the hardware is already at the Dallas datacenter, it's just waiting on the cages and racks to be built 17:30 -!- tmus [~tmus@cpe.atm2-0-1271101.0x50a0a16e.taanxx1.customer.tele.dk] has joined #linode 17:30 < afv-13> it's just the oh so efficient TP guys they're waiting on 17:30 < tierra> but yeah, I guess that last post sums it up 17:31 < iggy> atlanta is the one with blocked ports... I'd never move there (even though I don't use any of those ports personally) 17:32 < rk> tierra: I'm not completely sure what that post sums up. Does it concludt that it would take another one to two months, as the cages is not built? Or doest it suggest that the upgrade is just around the corner? 17:32 * tierra is also patiently awaiting upgrades at Dallas too btw 17:32 < rk> Many are I guess... 17:33 < caker> rk: they're telling us cage will be ready by Friday, then they have to install our hardware (they got it all last week) 17:33 < iggy> I think it was mentioned earlier today in here that the cages will be done Friday 17:33 < tierra> well, according to what TP told caker, they should have been up already, I don't honestly think it'll drag out any more than another week 17:33 < caker> so my guess is worst case about another week +/- a few days 17:33 < caker> after they rack stuff, we're ready to rock 17:34 < caker> all the boxes are configured 17:34 < iggy> I would say mention that in the thread, but as soon as you do, tp will fuck something else up 17:34 < rk> Okay, another week sounds worth waiting for. ;o) 17:35 -!- wastrel [~wastrel@76.8.67.2] has quit [Quit: bye now] 17:35 < rk> Thanks. 17:39 < tmus> hi there - any plans to create a CentOS 5 image? 17:40 < caker> tmus: yes. mikegrb is working on that 17:40 < caker> mikegrb: how's CentOS 5 going? 17:41 < tmus> caker: aah, sounds great that it's not just being ignored .) 17:42 -!- tibbetts [~tibbetts@onat.streambase.com] has quit [Ping timeout: 480 seconds] 17:43 < afv-13> caker: has there ever been demand for a rhel image? 17:43 -!- tmus [~tmus@cpe.atm2-0-1271101.0x50a0a16e.taanxx1.customer.tele.dk] has quit [Quit: I wonder what this switch is for...] 17:43 < caker> afv-13: CentOS, yes. 17:43 -!- tmus [~tmus@cpe.atm2-0-1271101.0x50a0a16e.taanxx1.customer.tele.dk] has joined #linode 17:43 < afv-13> but not rhel itself? 17:44 < caker> rhel implies their kernel and money to them 17:44 < afv-13> oh of course 17:45 < afv-13> other than the money there must be some kind of kernel workaround 17:45 -!- alinrosca [~alinrosca@adsl-75-33-74-134.dsl.bcvloh.sbcglobal.net] has joined #linode 17:45 < afv-13> restricted to those who paid for the licence 17:45 < kvandivo> caker: lessons learned wrt announcing upgrades before you have the hardware in place? 17:45 < tierra> which is usually only useful for those running exotic hardware they need support for, and that's not the case on a Linode ;) 17:45 < iggy> you use linode kernels, so it wouldn't matter 17:45 < tmus> afv-13: for stuff such as this, the open repos for updates and such is a great reason for centos over redha 17:45 < tmus> ...t 17:46 < afv-13> i'm not a fan of rhel, just curious 17:46 -!- clanehin [~clanehin@69.134.236.134] has quit [Remote host closed the connection] 17:47 < tmus> :) I like the redhat family of distros a lot, rarely use RHEL outside of work for the reasons mentioned, though 17:47 < iggy> I'm kind of wondering what would be the point? 17:47 < tmus> centos or fedora for all "the other stuff" 17:47 < iggy> redhat wouldn't support you on linode anyways 17:49 < tmus> anyways, glad to hear it's coming (centos) - catch you all later - itøs getting late here 17:49 < tmus> see ya 17:49 < tmus> :) 17:49 < afv-13> no customizations to anything or loose your id 17:49 -!- tmus [~tmus@cpe.atm2-0-1271101.0x50a0a16e.taanxx1.customer.tele.dk] has quit [Quit: I wonder what this switch is for...] 17:50 -!- TJF [~MirandaIM@pat.foulston.com] has quit [Quit: buh bye] 17:50 < caker> kvandivo: not really. I anticipated some of this, but it was an educated gamble (benefits of getting the word out sooner and risk customer weirdness with some upgraded and others not, vs waiting and losing more business, etc)... Nothing with my previous experiences woud have made me expect a delay such as this 17:50 < caker> ... with TP 17:52 -!- memset0x90 [~memset0x9@april.kalieda.net] has joined #linode 17:52 < caker> if we took the TP delays out of the scenario, everyone would have been upgraded ... (/me looks at Tom's tpdelay schedule) about three weeks ago 17:54 < encode> lovely 17:55 < encode> i liked knowing ahead of schedule - it upgrading was so vital there was always atlanta 17:56 < encode> s/it/if/ 17:57 < alinrosca> We recently bought vps hosting from Linode, and configured it to transfer our web site from our current vps host to Linode. However, the site we uploaded is loading very, very slow. We're using teh same operating system as with the previous host (Fedora 6); on teh previous host, the site is loading much faster. 17:58 < alinrosca> any ideas why? 18:03 -!- alinrosca [~alinrosca@adsl-75-33-74-134.dsl.bcvloh.sbcglobal.net] has quit [Quit: alinrosca] 18:03 -!- alinrosca [~alinrosca@adsl-75-33-74-134.dsl.bcvloh.sbcglobal.net] has joined #linode 18:05 < alinrosca> dWe recently bought vps hosting from Linode, and configured it to transfer our web site from our current vps host to Linode. However, the site we uploaded is loading very, very slow. We're using teh same operating system as with the previous host (Fedora 6); on teh previous host, the site is loading much faster. Any ideas why? 18:06 < Eman> does it rely heavily on an sql database? 18:06 -!- futhin [~wunderwan@bespin.org] has joined #linode 18:06 < futhin> mikegrb caker, you guys know about slicehost right? 18:06 < alinrosca> yes 18:07 < alinrosca> but it's the same application as on the old host 18:07 < futhin> i find it rather ironic that linode has like 200 slots open and slicehost has none, considering that that slicehost came around after linode 18:07 < futhin> slicehost & linode are very similar imo 18:07 < futhin> in terms of reputation & irc channel, etc 18:08 < Eman> alinrosca: iirc, the linode wiki has some tuning guides for apache and mysql 18:08 < futhin> main difference is that slicehost is part of the ruby on rails community, and the irc channel is in freenode 18:09 < encode> i've never heard of slicehost 18:09 < Eman> same here 18:09 < encode> and i did a lot of searching before i came to linode 18:09 < encode> so maybe slicehost are less than 2 years old? 18:10 < futhin> slicehost is recent, in the last 6 months 18:10 -!- kevincody [nobody@c-24-147-111-136.hsd1.ma.comcast.net] has joined #linode 18:10 < encode> ouch 18:10 < encode> i dont think i could rely on a company that new 18:11 < futhin> yeah linode most likely has a way bigger customer base 18:11 < encode> they can't have experienced any major disasters yet 18:11 < futhin> has linode? 18:11 < futhin> good point though 18:12 < encode> well, they've had power outages at datacentres 18:12 < encode> that were unscheduled 18:12 < encode> which i count as a pretty major disaster 18:12 < encode> losing power to all of your servers at a DC is never a good thing 18:12 < kevincody> actually it's a massively bad thing 18:12 < encode> well, yeah 18:12 < encode> but linode handled it well 18:13 < encode> and theres never been anything in linode's control that has caused a major outage 18:13 < futhin> whats the status of Xen wrt linode? 18:13 < encode> considering they've been around 5 years or so, thats a pretty good indicator of their expertise 18:13 < kevincody> yes, that is good 18:13 < encode> compared to the relatively new company 6 months old 18:14 < encode> futhin: still in beta afaik 18:14 < kevincody> depending on price i'd get a server from a 6 month old company, but only as a node in a cluster - not as a standalone 18:14 < encode> i only know of the one Xen host that is used by the beta 18:14 < kevincody> i'd love to have the ease of all my nodes being linodes, but that's as bad as a datacenter with only one power feed... 18:15 < futhin> how can you tell how much bandwidth you've used up? does the linode acct show that clearly ? 18:15 < Deckert> futhin: Linode has so many nodes open due to recent massive upgrades to their infrastructure. 18:15 < encode> and im pretty sure there are outstanding issues that need resolving before production 18:15 < kevincody> futhin, yes that's clear as day on the web interface 18:15 < Deckert> futhin: have a look at the demo dashboard 18:15 < encode> futhin: theres a 24hr and a weekly graph of bandwidth 18:15 < encode> as well as a total incoming and outgoing 18:15 < encode> plus averages on the graphs 18:15 < futhin> cool 18:16 < mikegrb> encode: yeah, we likesly won't be going with xen due to massively reduced performance 18:16 < kevincody> what would xen buy, anyway, besides LKM's 18:16 < mikegrb> encode: 24hr and monthly graph ;) 18:16 < futhin> mikegrb: what? i thought xen had better performance 18:16 < encode> mikegrb: oh, i was close 18:16 < mikegrb> futhin: in theory 18:16 < Deckert> kevincody: it would buy reduced control over IO and a less friendly dashboard :-) 18:16 < mikegrb> futhin: it has less cpu over head 18:17 < mikegrb> futhin: in practice is another issue 18:17 -!- megaversal [~455d7f80@webuser.linode.com] has quit [Quit: CGI:IRC (Session timeout)] 18:17 < futhin> interesting 18:17 < encode> KVM in the kernel should help uml even more, shouldn't it? 18:17 < mikegrb> http://thegrebs.com/~michael/sshot/dashboard.png <-- the dashboard 18:17 < mikegrb> encode: yes, would make uml cpu overhead on par with xen 18:17 < alinrosca> our admin is telling me we looked over the tuning guides for apache and mysql on linode wiki and our settings seem ok. 18:18 < encode> alinrosca: whats your swap usage like? 18:18 < futhin> mikegrb: i had a linode before but i wasn't happy with the latency so i upgraded to a dedicated server, but then i dropped it for dreamhost heh, now i'm ready for a vps again :P 18:18 -!- clanehin [~clanehin@user-0c99956.cable.mindspring.com] has joined #linode 18:18 < encode> alinrosca: and what is the status of your io tokens? 18:20 < Deckert> alinrosca: and maybe check out the load on your host, just to be sure 18:20 < futhin> the latency could've been a fluke though.. 18:20 < encode> i've never had any latency issues 18:20 < encode> depends on what you're running I guess 18:21 < kevincody> what were you running that latency became an issue? 18:21 < futhin> well i was running apache + mysql 18:21 < caker> ha ha slicehost :-p 18:21 -!- plamarao [~pedro.lam@201.19.123.249] has joined #linode 18:21 < futhin> but i was experiencing latency in terms of sshing in 18:21 < futhin> like it would take 10 seconds to ssh in 18:22 < Bdragon> dns lookup? 18:22 < encode> futhin: ever try sshing with -v to see what was going on in that 10secs? 18:22 < Bdragon> Check resolv.conf.... 18:22 < futhin> nah, i use sitelutions for dns 18:22 < futhin> encode: nope 18:22 < Bdragon> No, the ssh daemon looking YOU up 18:22 < encode> caker: whats funny about slicehost? given I know nothing about it other than what was just mentioned 18:23 < futhin> Bdragon: ah, possible 18:23 < encode> or are you just commenting on the laughable comparison 18:23 < caker> encode: I was just being an ass. They've got a lot of buzz lately, but don't have the resources to continue purchasing new equipment. Hence, they now require at least 3 months pre-payment to even get into their signup queue, and even then, you've gotta wait weeks 18:23 < Bdragon> Cuz the 10 seconds COULD be a reverse lookup timeout 18:25 < caker> I'm still amazed how uneducated their clients seem to be, with regard to slicehost alternatives (namely Linode) -- with features they're just rolling out getting covered in blogs, etc, that we've had for years 18:25 < caker> something I'm working on correcting :) 18:26 < caker> I will say, slicehost is one of the main reasons I wanted to push out our upgrades sooner rather than later. 18:27 < kevincody> caker, by the way: i have arla working 18:27 < futhin> caker: yeah, linode is definitely undermarketed to the rails community and in general.. 18:27 < caker> And as far as them going with xen (if you want to call that an advantage), it won't make a difference by the end of this year is my guess 18:27 < caker> kevincody: kick ass 18:27 < futhin> caker: p.s. i think it would help ALOT if you and mikegrb setup a channel in freenode 18:27 < encode> arla is for afs? 18:27 < caker> kevincody: how difficult do you foresee maintaining those patches for newer kernel releases? 18:27 < kevincody> encode, yes, one of two decent available *nix client sides 18:28 < encode> kevincody: are you going to post a howto on the wiki? 18:28 < futhin> caker: freenode is the irc network to have a channel on if you're serious about linode.. not asking you to move the channel over, but to setup a presence in freenode 18:28 < caker> I guess I could do the supybot relay thing with linbot 18:28 < kevincody> caker, difficult to say - they were a straight-up "cp ~/arla-0.90/nnpfs/linux/*.c /usr/src/linux/fs/nnpfs" 18:28 < futhin> caker: a relay isn't as ideal as an actual presence 18:29 < encode> maybe just create a channel, set the topic to redirect people here, and +m the channel 18:30 * Bdragon thinks that's kinda iffy... Better check freenode AUP first 18:31 < futhin> caker: there's already 54 people in #slicehost on freenode 18:32 -!- tehdan [~dan@omicronpersi8.plus.com] has joined #linode 18:32 < futhin> and the channel is only 13 weeks old 18:32 < caker> cool 18:33 < kevincody> caker, the arla-0.90 sources in /afs/stacken.kth.se/ftp/pub/arla/ are dated Jan 10 of this year - my guess is it will remain stable until kernel.org decides to screw with the vfs layer, which probably isn't likely 18:33 < futhin> caker: mikegrb: please come to #linode on freenode, not asking you to move or anything, just setup another channel.. it will be worth it, it'll become bigger than this one in a few months, i swear 18:33 < Bdragon> ( http://freenode.net/policy.shtml#advertising paragraph III) 18:34 < alinrosca> : swap usage: zero (we're using 160 out of our 256M); what's io tokens? what do you mean by load on our host - traffic or hard disk? 18:34 < encode> alinrosca: just load 18:34 < Bdragon> alinrosca: cat /proc/io_status 18:34 < encode> run uptime 18:34 < encode> see what the load averages are 18:34 < encode> and then do what Bdragon said 18:35 < Bdragon> also, vmstat 5 18:35 < Bdragon> high numbers in the last column "wa" are bad... 18:36 < encode> if i recall correctly, wa isn't shown in a 2.4 kernel 18:36 < encode> alinrosca: which distro, what kernel? 18:36 < kevincody> encode, suppose i should put it up on the wiki 18:37 < encode> kevincody: just a suggestion...you're under no obligation of course 18:37 < kevincody> do you refer to the userspace setup or the kernel patch maint, by the way? 18:37 < mikegrb> si and so are two of the most important columns on a linode 18:37 < mikegrb> they should be 0 on most all lines 18:37 < kevincody> encode, i think i'm under at least a token obligation, seeing how caker has to build all the kernels 18:38 < kevincody> but it's also an ad hoc build right now so let me clean it up some 18:39 < alinrosca> wa = 0. 18:39 < encode> i don't personally see a need for afs for my linode, but im sure someone else will 18:40 < encode> and it seems so much better than nfs 18:40 < alinrosca> vmstat procs -----------memory---------- ---swap-- -----io---- --system-- -----cpu------r b swpd free buff cache si so bi bo in cs us sy id wa st 0 0 0 87276 6528 49576 0 0 22 15 139 37 1 0 99 0 0 18:40 < kevincody> for wide area sharing yes, much better than nfs 18:40 < alinrosca> uptime 18:39:47 up 50 min, 1 user, load average: 0.49, 0.50, 0.45 18:41 < kevincody> it's not really into its own though until you've got several geographically diverse servers involved 18:41 < kevincody> it's REALLY SWEET having a cluster-of-shell system spanning the continent ;) 18:41 < alinrosca> fedora core 6, 2.6 kernel 18:41 < mikegrb> if you've only been up 50 minutes and already have so much in swap, I'd look there 18:43 -!- Smark [~Smark@cm-24-121-61-66.payson.az.npgco.com] has joined #linode 18:44 < alinrosca> swap: si=0, so=0 (sorry for the formatting above) 18:45 < encode> kevincody: forgivemy ignorance, but what is a cluster-of-shell system? 18:46 < kevincody> several shell servers with a common username/password database and shared home directories 18:46 < encode> ahh 18:46 < kevincody> most often done with ldap and nfs these days 18:47 < kevincody> thus they're almost always local-area systems 18:47 < encode> that does sound cool 18:47 < encode> for latency, load balancing and availability 18:48 < kevincody> exactly. keep your i/o patterns in mind, but it's how you scale web servers past the load a physical box can take, as well 18:49 < encode> if only i had a website that required that much performance 18:49 < kevincody> doing it with openafs, opens up the possibility of geographic redundancy - either so you can start routing clients to the nearest-server via smart dns answers, or because you're worried about your data getting nuked 18:49 < kevincody> doing it for performance, usually you'd want to go SAN, unless your web server load is seriously way way more reads than writes 18:49 * encode wonders if akamai uses something similar 18:50 < kevincody> they are making the decisions during the dns lookup, yes 18:50 < kevincody> but they're using a proprietary distribution system 18:50 < encode> yeah, but afs etc too 18:50 < kevincody> their system deprives even the network admins of knowing what's being stored in their system 18:51 < kevincody> all the names mangled 18:51 < encode> heh 18:51 < alinrosca> our buff and cache at Linode are 6528 and 49576, while at the old host they're 0. any idea what that means? where should we look to fix this? 18:51 < alinrosca> thanks 18:52 < kevincody> actually something seems to be up, my linode is doing -NOTHING- and taking up a .5 load 18:52 < tierra> caker: you have any idea how SH handles disk i/o? 18:52 < caker> tierra: they don't :) 18:52 < tierra> heh 18:52 < encode> oooh, nasty 18:53 < caker> they also assign nodes ad-hoc to servers, so you may be the one biggie on a box with lots of smaller, thrashing ones 18:53 < tierra> I can see that biting them in the ass real quick 18:53 < Eman> i just migrated from host9 to host74... much much much better :D 18:53 < encode> thats two really big reasons why linode is Real Good (TM) 18:55 * encode joins slicehost, with his hostmask as the default linode rdns 18:56 < caker> Eman: cool! 18:56 < tasaro> Eman: I thought we had to pry host9 from your dying hands? 18:57 < Eman> lol 18:57 < mikegrb> lolz 18:57 -!- alinrosca [~alinrosca@adsl-75-33-74-134.dsl.bcvloh.sbcglobal.net] has left #linode [] 18:57 < caker> kevincody: I think that's an artifact from a cosmetic thing in recent 2.6-um kernels... haven't been able to track it down yet (haven't looked all that hard, either) 18:57 < Eman> its stable, but its just too damn slow now 18:57 < caker> either that, or it's doing something weird with threads or process accounting 18:58 < tag> hurm 18:58 < tag> host 42 seems to generally have a pretty high load these days 18:58 < mikegrb> tag: I was just looking at t hat actually 18:58 < kevincody> which means my write tests last night were pushing 700 kbit/s at about an 0.15, i like that much better 18:59 < tag> so was I because it's the box my shit is on :-( 19:00 < tierra> I would find it odd that you would be looking at it when your node isn't on that host... :) 19:00 < tag> my node has used 1% of the CPU time this month :-) 19:01 < mikegrb> tag: should start getting better 19:01 < tag> mikegrb: sweet, what happened? 19:01 < caker> thrasher, most likely 19:01 < tag> "thrasher" ? 19:02 < caker> can't wait until everyone's on new hardware... The loadaverages for the new hosts doesn't even register on our "Top 10 host load" meter thing 19:02 < caker> tag: someone thrashing the disks, mostly likely because they're swapping" 19:02 < tag> yeah 19:03 < kevincody> tag, which in turn means, someone that needs to buy more memory that isn't 19:03 < kevincody> or reconfigure their apps 19:03 < tierra> I'm on host 43, so I'm guessing that means 42 is in Dallas with 43, and is one of the hosts pending replacement with brand spankin' new hardware in the next week 19:03 < tag> kevincody: no kidding. 19:03 < tag> sweet 19:03 < tierra> which also means your friendly thashing neighbors will be getting more ram too 19:04 < tag> I'll be looking forward to taht 19:04 < tierra> heh 19:04 < tag> that, even. 19:05 * tag crosses his fingers and prays 19:06 < tag> I hate doing remote upgrades. 19:11 < mikegrb> http://thegrebs.com/~michael/sshot/lnadmin.png 19:11 < mikegrb> tag: ^ 19:11 < Eman> im thinking of switching to lighttpd, are there any prebuilt packages available? (centos) 19:11 -!- pygi [~mario@78-0-18-10.adsl.net.t-com.hr] has joined #linode 19:11 < caker> pygi: hola 19:11 < pygi> hi caker :) 19:11 < tag> wow that's a cute UI 19:11 < caker> pygi: how's it going? 19:12 < pygi> caker, pretty good, a bit tired ^_^ low-level C is bugging me :) 19:12 < pygi> caker, what about you? 19:12 < caker> good good. what are you hacking on? 19:12 < pygi> one of the crucial linux components =) 19:12 < pygi> libburnia project =) 19:13 < encode> ooh low level C 19:14 < encode> so much fun 19:14 < encode> so easy to screw things up 19:14 * encode has fond memories of the OS subject at uni, modifying minix 19:14 < pygi> encode, it's fun ^_^ 19:14 < encode> yeah, unfortunately i dont have time / reason to do any these days 19:15 * tag has fond memories of C, but no memories of university 19:15 < encode> too much partying? 19:15 < pygi> encode, I do ... the need to fix linux in the area which is covered by the libburnia project, including the same area in the linux kernel :) 19:15 * encode looks up libburnia 19:15 < tag> no, I didn't go 19:16 < tag> :-D 19:16 < encode> tag: i see...nothing wrong with that 19:16 < tag> nope 19:16 < ajmitch_> hey pygi 19:16 < pygi> haha, ajmitch_ is here :) 19:16 < ajmitch_> of course 19:17 < pygi> I gotta run now or else ajmitch_ will eat me :( 19:17 < encode> i really enjoyed uni though - gave me a different, perhaps even more mature, perspective on things 19:17 < ajmitch_> >:) 19:17 < pygi> you're everywhere ajmitch_ ? 19:17 < ajmitch_> yep 19:17 < pygi> or you're just stalking me? 19:17 < ajmitch_> nah, I don't have time to stalk you 19:17 < ajmitch_> sorry 19:17 < pygi> but you do have time to be everywhere? 19:17 < pygi> that's contradictive :P 19:17 < pygi> anyway, /me is back to hacking and fixing stuff =) 19:18 < ajmitch_> heh, bye 19:18 < pygi> I've got SoC to do :) 19:18 * ajmitch_ has other hacking to do 19:18 < ajmitch_> including stuff I did for SoC last year 19:18 < pygi> well, my SoC = my usual work :P 19:18 < ajmitch_> still bugs 19:18 < pygi> ajmitch_, yea, well :-/ 19:18 -!- tag [~smccoy@peer.sef.marchex.com] has quit [Quit: Lost terminal] 19:18 < ajmitch_> & still so much more I can do :) 19:19 < pygi> and you thought you'd do it in three months ;) 19:19 * pygi hides =) 19:19 < ajmitch_> the essential client stuff was :P 19:19 < pygi> haha 19:19 < ajmitch_> excluding some minor breakage 19:20 < pygi> minor = xD 19:20 * pygi hides=) 19:20 < ajmitch_> you're so helpful 19:20 < pygi> heh, do I see a sarcasm here :-/ 19:20 < ajmitch_> makes me wonder why I bother hacking on it at all :P 19:20 < pygi> sorry, I'm just joking ... it's 1:20AM 19:21 < pygi> seems you can't take jokes :( 19:21 < ajmitch_> nope 19:21 < pygi> oh well, laters then 19:21 -!- pygi [~mario@78-0-18-10.adsl.net.t-com.hr] has left #linode [Leaving] 19:22 -!- tag [~smccoy@peer.sef.marchex.com] has joined #linode 19:22 < tag> see, that's why I don't work in operations 19:22 < tag> I just looked at a terminal, carefully checked to make sure apt was finished running...decided it was time to restart the box, and typed sudo init 6 -- turns out I didn't look at the host part of my prompt (or my title bar for that matter) and restarted my workstation 19:24 < tag> it was supposed to be my linode I was restarting :-) 19:28 < rk> tag: It is worst when you do it the other way around. 19:29 < tag> yes, I know...I've done that before 19:29 < tag> only it was even worse - it was init 0 19:29 < tag> so I had to call up the noc and have them walk in and power our rack back up 19:29 < rk> do'h, not good 19:31 < rk> I found changing the prompt (possibly only the root prompt) on you servers to something like dark red or magenta or anything else out of the usual helps. 19:31 < tag> caker: hey, when you move host42, will it have more ram available? 19:32 < tag> yeah I should set up my zshrc to make my linode prompt be some color 19:34 * Bdragon has a modified phil's on some of his stuff 19:34 < tag> Additional RAM 0 MB 19:34 < tag> :-( 19:35 < rk> Also a * dns host can be fatal. Trying to connect to a development server, mispelling its hostname and instead ending up on a production server and only realizing after doing some experimental changes. 19:36 < Bdragon> 3 line version of this (have jobs and procs on middle line..) ---> http://www.aperiodic.net/phil/prompt/ 19:39 < rk> Bdragon: looks cool, but it still has the same problem if you use it on all your hosts. 19:39 < Bdragon> Only use it in a few places, I'm on dialup so I tend to stick with a really light prompt most of the time 19:40 < rk> I didn't know anyone was still on dialup. 19:41 < tag> amazing isn't it 19:41 < Bdragon> Apr 17 17:28:59 dialer chat[5137]: CONNECT 26400/ARQ/V34/LAPM/V42BIS 19:41 < tag> Bdragon: I use a one line no color prompt....because, I don't need my terminal to look like an old BBS :-) 19:41 < Bdragon> Rural South Dakota 19:41 < tag> 26400, crummy 19:41 < tag> Bdragon: you have cellular networks out there? 19:42 < Bdragon> I'm on a cell edge, but it's analog for a few miles 19:42 < tag> crummy, so wimax is out of the question 19:42 < tag> why do you live there? 19:42 < tag> :-) 19:43 < Bdragon> It's home. *shrug* 19:43 < tag> to each their own. 19:43 < Bdragon> I can always drive to town with a stack of cds when I need something faster ;) 19:43 * tag prefers smack in the middle of a major metropolis. 19:44 < tag> I use wimax though 'cause I don't have cable/dsl/etc available (I live on a sailboat) 19:44 < mikegrb> mmm living on a sailboat 19:45 < Bdragon> Last mile' 19:45 < Bdragon> 's a bitch 19:45 < mikegrb> tag: this one? http://thegrebs.com/~michael/myboat.jpg 19:45 < tag> yeah it's pretty fun. Plus the wimax works when I'm about 2 miles off shore, makes for some nice unplugged datturday afternoon drifts 19:45 < tag> no, a sailboat. That's a motor yacht 19:46 < mikegrb> ;) 19:46 < tag> http://dav.blisted.org/scott/boat/ 19:46 < mikegrb> oh nice 19:46 < tag> that's my boat, those photos are a little dated though - I took out that ugly TV/VCR and put a 25" TFT in it's place, few other things. 19:47 < tag> little bit of damage on hull from a major windstorm last winter :-( will be easy to fix though next time I haul it out 19:48 -!- tibbetts [~tibbetts@c-24-34-111-180.hsd1.ma.comcast.net] has joined #linode 19:48 < tag> This one is quite close to where I keep 'er parked: http://dav.blisted.org/scott/boat/IMG_7783.JPG 19:54 < caker> tag: yes -- you'll get the doubling of resources, and there'll still be headroom on the host 19:55 < caker> tag: damn fine boat 19:55 < mikegrb> caker: yes 19:55 < mikegrb> caker: let's steal it 19:55 < encode> mikegrb: you can't 19:55 < encode> im already stealing it 19:55 < mikegrb> :< 19:55 < encode> we can't both steal it 19:56 < mikegrb> encode: I'll share 19:56 < encode> ok 19:56 < encode> thats good, because i haven' 19:56 < encode> t a clue how to use a sailboat 19:56 * mikegrb still wants this one http://thegrebs.com/~michael/myboat.jpg 19:56 < mikegrb> it never goes anywhere, it's always in the same spot 19:56 < mikegrb> they don't use it, won't notice it missing! 19:56 < caker> Mikegrb gave Mona Lisa that smile. 19:57 < caker> Mikegrb has two speeds: walk and kill. 19:57 < mikegrb> or this one http://thegrebs.com/~michael/stad_amsterdam-004.jpg 19:57 < encode> that myboat is a very nice looking boat 19:57 < encode> i'll take two 19:57 < encode> (need a hot spare) 19:57 < encode> that was also a very lame pun 19:57 < encode> for which i shall now apologise 20:02 < tierra> mikegrb: you manage Linodes with a Mac? That's depressing... heh 20:03 < encode> tierra: hrm, what prompted that comment? did you only just realise mikegrb was a mac fan? 20:03 < tierra> 17:11 <@mikegrb> http://thegrebs.com/~michael/sshot/lnadmin.png 20:03 < caker> Superman owns a pair of Mikegrb pajamas. 20:04 < mikegrb> caker: he so does 20:04 < tierra> I know he's mentioned having a Mac before... I didn't realize he wrote his management tools for it though 20:04 < mikegrb> tierra: it runs on linux too! 20:04 < tierra> out of curiosity then, what did you write it with? 20:04 * mikegrb whispers realbasic 20:05 < tierra> hehe 20:05 < mikegrb> http://www.realsoftware.com/ 20:05 < mikegrb> http://thegrebs.com/~michael/sshot/about.png 20:06 < encode> that looks cool 20:06 < tag> thanks 20:06 < tag> You all can have luck stealing it 20:06 < mikegrb> handles activations too 20:06 < mikegrb> and growl/speech notification of new signups 20:06 < encode> mikegrb: but can it make coffee? 20:06 < mikegrb> encode: yes 20:06 < mikegrb> uses the x10 stuffs 20:07 < encode> oh yeah, growl. i haven't tried your growl script yet 20:07 < tag> heh 20:07 < tag> that linode admin thing is cute, probably was pretty fun to build 20:08 * tierra can't say he'd enjoy working in BASIC 20:10 < tag> it's written in basic? 20:10 < opello> Bdragon: rural south dakota, eh :) where abouts? 20:10 < tag> that sucks 20:10 < mikegrb> tierra: well it's not really BASIC 20:10 < mikegrb> tierra: it's all OO stuff 20:10 < Bdragon> Newton Hills area 20:10 < caker> mikegrb: is it OO? 20:10 < caker> nod 20:10 < tierra> right, and neither is Visual BAsic 20:11 < tierra> *Basic 20:11 < tag> that makes for multiple "interesting" technology selections at linode 20:11 < opello> Bdragon: heh, i'm a bit north of you :) brandon 20:11 < Bdragon> lol 20:11 < mikegrb> lolz 20:11 < opello> but i have mbit dsl 20:12 < Bdragon> Yeah, but I'm 7 miles out of Canton, so... 20:12 < opello> heh yeah :) 20:13 < Bdragon> Small world XD 20:13 < opello> heh, indeed 20:14 < opello> fun though :) 20:57 < kevincody> caker, do you have an eta. on the specific capability of having more than one linode to a customer? 20:58 < Smark> there an easy way to see the overall disk usage you are using (from the shell) 20:58 < Bdragon> df -h? 20:59 < Smark> ty 21:14 -!- Dreamr_3 [~Dreamer3@74-129-108-228.dhcp.insightbb.com] has joined #linode 21:30 < caker> kevincody: EOM or so for just the basic multiple Linodes under one customer and ability to switch between them w/o logging out/in stuff 21:30 < kevincody> cool, that's about right i think... 21:31 < kevincody> can dallas in any way be persuaded to open up ports for me? 21:31 < caker> dallas is open / noone has ever complained 21:31 < kevincody> oh ok, the website says they're locked down out the wazoo 21:31 < caker> Atlanta, if it's non irc I think there's still a chance in hell; if for IRC no 21:32 < caker> which dallas ports did you need poked through? 21:32 < kevincody> oh it's atlanta that's bolted up. hmm. well, atlanta's the closest to home, easiest to do without anyway... 21:33 < kevincody> 700{1,2,3,5,7}; 7021 21:33 < kevincody> all udp 21:34 < caker> bah .. I have a feeling the'd be reluctant since it's so close to irc ports 21:34 < caker> but I can try 21:34 < kevincody> irc is tcp 21:34 < caker> yeah, I didn't look up after the first line 21:35 < caker> what justification / application shall I tell them to help convince? 21:35 < kevincody> all this hinging on my proof of concept and short term beta going well, of course 21:35 < caker> both outgoing/incoming, I assume? 21:35 < kevincody> so i can keep all 3 of my servers as linodes, running openafs 21:37 < kevincody> the "product" i'll be selling, if you want to call it that, is having your website hosted in 3 places around the country, plus it's easy to securely update any of them from your laptop 21:37 < caker> ok, so those ports are for your custom app, not for openafs specifically? 21:37 < kevincody> the udp is for openafs specifically 21:37 < caker> ok 21:37 < kevincody> the tcp are standard ports 21:37 < kevincody> and will have the expected apps listening on them 21:38 < kevincody> also, 123 udp for ntp in both directions; 500/4500 udp and ip/esp for ipsec 21:38 < caker> 123 udp isnt blocked, is it? 21:38 < kevincody> no idea i'm not in atlanta 21:38 < caker> ok 21:38 < kevincody> i'm trying to do standard stuff in a bleeding edge way - GSSAPI auth on smtp/imaps, for example 21:39 < kevincody> geekish example; imagine being able to set your mail reader to unix Maildir access, and see the same spool from anywhere in the world, with automatic 3 location replication that you don't even have to see 21:40 < caker> this whole replication / global fs thing is a turn on, no doubt. 21:40 < kevincody> ;) tell ya, it was a trip issuing a shell command to trigger a cross country replica of a file that had been saved with vim running on my local cpu. 21:45 < tierra> honestly, I wouldn't normally think Linode would be that great of host to go with for that kind of thing... I've always seen Linode as low bandwidth and storage capacities traded off for quality of service and stability, and it sounds like this OpenAFS stuff could result in high bandwidth and storage requirements 21:47 < tierra> 1GB xfer on one client would be equal to 4GB (or so) overall bandwidth used between 3 servers hosting afs copies right? 21:47 < kevincody> not really on the storage requirements; replication is done in shots 21:48 < kevincody> if that was 1gb of writes yes; you wouldn't use this for live distribution, except small time stuff like say a unix home directory 21:48 < kevincody> the more likely uses is, from there things get downloaded from the nearest host 21:49 < kevincody> either using afs protocol and a client-local filesystem mount, or using a gateway - web server 21:49 < tierra> I can definately see the advantages to a distributed filesystem though. I'm assuming it works beautifully for a backup solution 21:49 < caker> until something eats your replicated file on one node 21:49 < kevincody> when the threat to be countered with a backup is loss of a whole facility, yes 21:49 < kevincody> only one node is read-write 21:50 < caker> ah 21:50 < kevincody> this has been around long enough to trust its basic semantics 21:50 < tierra> interesting 21:50 < kevincody> i believe ibm originally wrote it 21:51 < warewolf> two tablespoons of cinnamon 21:51 < kevincody> nodes can of course be promoted from readonly to read write 21:51 < warewolf> and two or three egg whites. 21:51 < warewolf> a half a stick of butter ... mmm .. melted. 21:51 < warewolf> stick it all in a bowl baby 21:52 < warewolf> stir it with a wooden spoon. 21:52 < warewolf> mix in a cup of flour, you'll be in heaven soon. 21:52 < kevincody> yeah the system is typically ibm'ish in design, multiple separate subcomponents under the control of its own "overseer" daemon, guess that was their approach to portability 22:01 -!- SpaceHobo [~spacehobo@host-84-9-49-210.bulldogdsl.com] has quit [Ping timeout: 480 seconds] 22:06 < kevincody> caker, what do you figure is the best burst bandwidth to Fremont? 22:07 < caker> from where / to where? 22:08 < kevincody> not sure where the user is; they just wondered if 1.1Mbit was the max, i told them no way but i have no idea -how- good it is 22:09 < caker> def not the max -- I've seen single source floods from across the globe saturate the link (not in pps, in mbit/sec) 22:09 < caker> if that counts :) 22:09 < kevincody> ;) 22:09 < kevincody> he looks like they're wired like a christmas tree 22:10 < caker> I can pull down about 7 or 8 mbit to my commiecast cable modem in Nashville, haven't tried from my parents' DSL here in NJ (visiting) 22:10 < encode> i've transferred files to/from my linode from other points in the US at around 6MB/s 22:12 < kevincody> in other words expect it to be every bit the first rate it's advertised to be. it'll be the customer, or several carriers' wide area fabric being the bottleneck 22:14 < encode> i'd say so 22:15 < encode> i think, although im not 100% sure, ive transferred files at 8MB/s sustained - but i couldn't guarantee that figure 22:15 < encode> i know ive done 6MB/s regularly 22:15 < encode> (thats megabytes btw) 22:18 < kevincody> that's pretty damn fast 22:18 < encode> yes 22:18 < encode> i was quite impressed 22:19 < encode> although my shell became less responsive during that period 22:19 < encode> no doubt due to lack of io tokens 22:21 < caker> encode: torrent, or http? 22:21 < caker> nm, stupid question 22:21 < caker> it was people seeding (multiple clients) that had issues 22:23 < kevincody> anyway, you wouldn't use afs for surfing, or spreading ripped dvd's around, or downloading pr0n 22:23 < encode> caker: dcc 22:24 < encode> irc file transfer 22:25 < caker> nod 22:27 < caker> kevincody: so the userspace tools for openafs/arla aren't pre-packaed by the popular distros? 22:27 < kevincody> can't speak for any other distro but gentoo 22:27 < kevincody> gentoo has openafs but not arla 22:33 < kevincody> arla required some work to compile the userland: krbafs library by source from mit, manually copy its include file into the arla build tree as a different filename, symlink the kerberosIV directory to Kerberos, remove a '.a' file left behind by the openafs build that caused a mislink 22:34 < kevincody> and actually only the daemon part of arla gets used; the "user" userland is the openafs binaries that come first in the system path - didn't function the other way around 22:35 < Bdragon> I thought milko was experimental. 22:37 < kevincody> it is 22:37 < kevincody> too experimental 22:38 < kevincody> i meant the client side daemon stuff, arlad 22:38 < Bdragon> Ahh 22:38 < kevincody> 'fs', 'vos', and whatever that the actual not-admin user calls, come from openafs 22:38 < Bdragon> OK, thought for a second you were planning on using milko ;) 22:38 < kevincody> hells no 22:38 < Bdragon> heh 22:38 < kevincody> the cell is already running 22:39 < kevincody> openafs server, arla is only on the linode so i can do secure web publishing, homedirs for shell accounts, etc. 22:42 -!- Bdragon is now known as Bdragon|zzz 22:45 -!- cnantais [~chad@d207-216-234-135.bchsia.telus.net] has joined #linode 22:46 < cnantais> what's a good mail server solution, just for mail forwarding, with the smallest memory footprint possible? 22:47 < kevincody> what kind of forwarding 22:47 < kevincody> where does the mail come from, local daemons or from port 25 22:47 < cnantais> just forwarding incoming email for addresses at my domain, off to a gmail account 22:48 < cnantais> 25 22:48 < kevincody> a relay 22:48 < efudd> http://www.gizmodo.com/gadgets/home-entertainment/synchronized-hd+dvd-purchase-by-fanboys-spike-amazon-rank-blu+ray-fanboys-set-to-retaliate-253131.php 22:49 < kevincody> you do NOT want a lightweight solution for that, it is a very serious thing to deploy a relay, due to the number of bozos paying people that got denied green cards to terrorize mail servers everywhere 22:50 < cnantais> kevincody: i see. 22:51 < Bdragon|zzz> Why not use "google apps for your domain"? 22:51 < cnantais> Bdragon|zzz: oh, good idea 23:02 -!- okokok [~Vampire@80-219-141-70.dclient.hispeed.ch] has quit [Quit: taisetsu to yuukoto wa subarashii tsukashi subarashii to yuukoto wa motto taisetsu desu] 23:15 -!- plamarao [~pedro.lam@201.19.123.249] has quit [Quit: Leaving] 23:15 -!- konoko [~Vampire@80-219-141-70.dclient.hispeed.ch] has joined #linode 23:30 -!- digitaljhelms [~digitaljh@ip68-1-24-216.pn.at.cox.net] has joined #linode 23:30 -!- afv-13 [~afv@rkdp-ip-nas-1-p120.telkom-ipnet.co.za] has quit [Remote host closed the connection] --- Log closed Tue Apr 17 23:59:01 2007