Back to Home / #linode / 2006 / 10 / Prev Day | Next Day
#linode IRC Logs for 2006-10-31

---Logopened Tue Oct 31 00:00:42 2006
00:01|-|GhostXz [] has joined #linode
00:02<Eman>windsor > kitchener
00:30|-|GhostXz [] has left #linode []
00:35|-|kokoko [] has quit [Ping timeout: 480 seconds]
00:41|-|kokoko [~Vampire@] has joined #linode
00:48|-|kokoko [~Vampire@] has quit [Quit: taisetsu to yuukoto wa subarashii tsukashi subarashii to yuukoto wa motto taisetsu desu]
00:50|-|gpd [~gpd@] has joined #linode
00:57|-|GhostXz [] has joined #linode
01:03|-|konoko [~Vampire@] has joined #linode
01:06|-|spr [] has quit [Quit: zz..spoon..zz]
01:06|-|GhostXz changed nick to GhostXz[FFXII]
01:42|-|londonguy [] has joined #linode
01:47<londonguy>I didn't realise I had to reboot to get my extra 20M of RAM haha.. I should read the forums more often!
01:53<londonguy>hmm I'm having trouble shutting it down :p tried root>shutdown -r now; then from lish going "shutdown" and from the webpage clicking "reboot".. now I got 3 jobs that are just in my queue on the webpage..
01:54|-|lucas [] has left #linode []
01:55<londonguy>shite I should have waited until it was daytime American time befoer rebooting but nooo I was too eager for that extra RAM lol
01:55<londonguy>hi Mike you're up super early
01:55<warewolf>he's not
01:55<warewolf>that's a triggered response
01:56<londonguy>actually I don't read the forums that often, largely because my node runs without a hitch. I had 207 days uptime before today.
01:56<londonguy>apart from the occasional external DOS hitting I've had no problems.. until yesterday when my host couldn't process i/o because one of the virtual servers had been compromised
01:57<londonguy>that was interesting.
01:57<warewolf>that free upgrade was quite some time ago
01:58<londonguy>I like linode because almost everything is limited - CPU, memory. But it seems the io_token bucket isn't perfect. I'll say this, though, in spite of packets being dropped and disk_io grinding down the server still hobbled along. Not happily, but it hobbled along.
01:59<londonguy>warewolf> I did wonder, as I saw a while ago that the new packages had extra RAM, and I just presumed it was for new accounts only. I've been happy with my 80 but can always use an extra 20!! I might even mod_perl my twiki now.
01:59<londonguy>my node is dead though can't seem to boot it.
01:59<warewolf>beware of mod_perl, it's quite a memory hog.
01:59<londonguy>exactly, warewolf. Which is why I've been running as a CGI app instead
02:00|-|GhostXz[FFXII] [] has left #linode []
02:00<londonguy>ok any ideas how I can almost cold reboot my virtual server.. :-S
02:00<AnMaster_>caker, or mikegrb here?
02:00<warewolf>do you have jobs stuck inqueue?
02:00<londonguy>yes, 3 of them
02:00<warewolf>AnMaster_: doubtful
02:00<londonguy>a shutdown, a reboot, and a shutdown
02:01<londonguy>what's wrong AnMaster_
02:01<warewolf>londonguy: give it 10 minutes, and then open a support ticket
02:01<AnMaster_>that they nullrouted our linode as it was being DDoSed and I wonder if the DDoS is over yet so it can be re-routed
02:01<AnMaster_>that is what is wrong
02:02<warewolf>AnMaster_: didja buy another IP?
02:02<AnMaster_>warewolf, tried to but got an error
02:02<warewolf>AnMaster_: how many IPs did you have before you tried to purchase another one?
02:04<londonguy>AnMaster_ what time did the nullroute start?
02:04<londonguy>warewolf I'm not going to raise a ticket; did that yesterday but it didn't get looked at till good ol US time anyway.
02:05<AnMaster>or around that
02:05<londonguy>warewolf and besides I just use the node for myself. If it was business I would be spending more anyway
02:05<AnMaster_>yep, had the scrollback on my other client ^
02:06<londonguy>AnMaster.. ok.. umm.. so about 10pm USA time.. you might have to wait until 3pm GMT+1 for the admins to wake up..
02:06<warewolf>londonguy: ah, well I don't know how linode monitors individual linodes themselves (as owners can shut down their linodes for an extended period of time) .. so ..
02:06<AnMaster_><badword /> <badword />
02:06<londonguy>warewolf after yesterday host24 might be a little sick anyway.
02:07<londonguy>AnMaster_ are you in Europe?
02:07<AnMaster_>I am
02:07<londonguy>I am in UK (as if you couldn't guess)
02:07<AnMaster_>somehow I could
02:07<fish1209>wonders what part of london londonguy is in
02:08<londonguy>I live in West London but work even further out west
02:08<fish1209>i have family in kensington
02:08<fish1209>i am from wales but been in usa for past 5 years
02:08<londonguy>how about you fish.. ah ok. Actually I bought my linode when I was in Australia. Personally I think it's the best deal world-wide (in English).. I see there are providers in Europe but their web pages are all in french or german or whatever.
02:09<londonguy>I'm thinking of working in the USA next sometime
02:09<londonguy>there aren't any good VPS resellers in UK yet I'm aware of.
02:11<londonguy>it's funny, I used to have an account with a web reseller - they were great, phone service, large disk space, fast transfers, loved it. But I can't let other companies own my domain names anymore. I have to run my own bind servers, need my own intelligence. So I don't think I'll ever go back to just a plain old web space provider again.
02:14<londonguy>actually I just had another thought
02:14<londonguy>you know how we have bandwidth and cpu graphs provided gratis (I love it)
02:15<londonguy>I wonder if linode would be interested in providing IO graphs as well so we can easily see how quickly we chew up our io tokens.
02:15|-|GhostXz[FFXII] [] has joined #linode
02:15<londonguy>I might think about running cacti on my own server
02:15<londonguy>and monitoring those things (cpu/memory/iotokens) for myself
02:15<londonguy>warewolf I think you're right I'm going to have to raise a ticket, that queue just isn't being processed.
02:29|-|andrew_j_w [] has joined #linode
02:33<londonguy>ah sweet!! Mikegrb just restarted the server daemon on host24!
02:34<@mikegrb>silly londonguy, I didn't /just/ do it, I did it 10 minutes ago ;)
02:35<@mikegrb>I was actually going back to see if you had issued another boot yet, was going to do that for you if you hadn't
02:40<londonguy>thanks sorry mikegrb just at work now and it's a personal linode, so came back to check on it afterwards. Didn't actually expect a response so soon, sorry to wake you!
02:40<warewolf>londonguy: there's a website I wrote (that needs to be updated) that will let you monitor your io tokens, and see where you stand in comparison to others:
02:42<londonguy>cool warewolf looking at it now
02:46<londonguy>quick question how do I abort a lish screen session
02:46<warewolf>^A d
02:46<warewolf>control A, then d
02:47<londonguy>love it! thanks
02:48<londonguy>hehe I have 92MB now ;) yay!!!
02:48<londonguy>I feel like a kid at Christmas
03:02|-|iamback [] has joined #linode
03:02|-|Turpan [] has quit [Read error: Connection reset by peer]
03:24<warewolf>"did you just say lol out loud?"
03:24<warewolf>"No, that would be lolol."
03:51|-|fish1209 [] has quit [Quit: [using [DeViL'z IRC] (build 3.21) by [ins4ne] from http://dIRC.RockZ.De]]
04:01|-|konoko [~Vampire@] has quit [Ping timeout: 480 seconds]
04:09|-|konoko [] has joined #linode
04:18|-|SpaceHobo [] has quit [Ping timeout: 480 seconds]
04:34|-|jekil [] has joined #linode
04:43|-|londonguy [] has quit [Quit: CGI:IRC (Ping timeout)]
04:45|-|jimcooncat [] has quit [Quit: Leaving]
04:50|-|jekil [] has quit [Ping timeout: 480 seconds]
05:06|-|SpaceHobo [~spacehobo@] has joined #linode
05:06|-|konoko [] has quit [Ping timeout: 480 seconds]
05:15|-|konoko [~Vampire@] has joined #linode
05:15|-|fish1209 [] has joined #linode
05:34<AnMaster_>mikegrb, is the DDoS over yet?
05:41<GhostXz[FFXII]>or caker
06:03|-|Shrapnal [] has quit [Quit: (I was using ) Version:(2.04) Wasted:(8 Hours 51 Minutes and 56 Seconds Online)]
06:26|-|linville [] has joined #linode
06:40|-|internat [] has joined #linode
07:03<AnMaster_>warewolf, "No, that would be lolol." <-- recent uf right?
07:04<AnMaster_>(a bit late for commenting what you said I know...)
07:04<warewolf>AnMaster_: ofcourse
07:05|-|GhostXz[FFXII] changed nick to GhostXz
07:24|-|internat [] has quit [Quit: This computer has gone to sleep]
07:56|-|NeonNero [] has joined #linode
08:22|-|konoko [~Vampire@] has quit [Ping timeout: 480 seconds]
08:42|-|sracatus [] has joined #linode
08:43|-|konoko [] has joined #linode
08:43<dc0e>anyone aware of host6 problems? getting no response from ssh backchannel or from my virtual server there
08:45<@caker>dc0e: looking
08:46|-|mauro_ [] has joined #linode
08:46<dc0e>caker thanks ssh -v says 'connection established' but so far doesn't get any further
08:46<@caker>looks like it's wedged
08:47<dc0e>caker: thanks; how to unwedget it?
08:47|-|sracatus [] has quit [Quit: CGI:IRC (Ping timeout)]
08:48<dc0e>ssh is making (slow) progress -- key exchange happening
08:50<dc0e>connection closed by host6; i'll wait...
08:51<mauro_>hi, i've read the topic..
08:51<mauro_>does anyone knows when the linode VS become up?
08:52<iggy>mauro_: rephrase the question
08:53|-|ElectricElf [] has quit [Remote host closed the connection]
08:53<mauro_>well =).. When the linode VS will work again?
08:53<@caker>one moment
08:53<mauro_>well =).. When will the linode VS work again?
08:53<mauro_>it' better
08:54<iggy>mauro_: not really, are you saying your linode has stopped responding?
08:55<mauro_>iggy: yes, it is dead
08:55<iggy>what host are you on?
08:55<mauro_>i can't connect with ssh
08:56<mauro_>i've tryed to restart the server but it is in queue
08:56<iggy>caker is working on it
08:56<mauro_>since 10 minutes
08:57<iggy>the first person in here to mention it was ~14mins ago
08:57<iggy>caker said he was looking into it then
08:57<iggy>about 4 mins ago, caker said one moment
08:58<iggy>if he had to reboot host6, then the linodes will come back up staggered
08:59<iggy>which usually takes anywhere from 5-15 mins for all of them to come back up depending on the number of linodes on a host and various other factors
09:01<mauro_>my linode is a mail server.. about 5 hours ago i've received some calls from my customers
09:01<mauro_>who said that the mail server doesn't work
09:02<SpaceHobo>the mail will queue
09:02<mauro_>but i've verified only now that the host6 has issues
09:03<mauro_>so i think the host6 has hang out about 4-5 hours ago, doesn't it?
09:03<@caker>no, it was working, just very slowly
09:03<linbot>New news from forums: Reboot: host6 in System and Network Status <>
09:04<mauro_>ah, ok
09:08<GhostXz>caker, is the attack done?
09:08<@caker>GhostXz: I'll ask
09:39|-|donnchaocaoimh [~donncha@] has joined #linode
09:42|-|NeonNero [] has quit [Quit: Everybody who believes in telekinesis, raise my hand!]
09:42|-|fish1209 [] has quit [Read error: Connection reset by peer]
09:46|-|ElectricElf [] has joined #linode
10:00|-|ElectricElf [] has quit [Remote host closed the connection]
10:07|-|Dreamr_3 [] has joined #linode
10:10|-|ElectricElf [] has joined #linode
10:13|-|Dreamr3 [] has quit [Ping timeout: 480 seconds]
10:15|-|FireSlash [~FireSlash@] has joined #linode
10:15|-|mauro_ [] has quit [Ping timeout: 480 seconds]
10:33|-|sracatus [] has joined #linode
10:38|-|gpd [~gpd@] has left #linode []
10:46|-|spr [] has joined #linode
10:55<linbot>New news from forums: availability linode 100 in Sales Questions and Answers <>
10:56<@caker>glue records are only required when the nameserver exists inside the domain itself (or a subdomain of that domain), correct?
10:56<@caker>I can't figure out why a dns manager beta tester is having issues pointing his domain's auth dns servers to ns[1|2] -- it's saying they need glue records
11:01|-|andrew_j_w [] has quit [Remote host closed the connection]
11:03|-|andrew_j_w [] has joined #linode
11:09|-|afv-13 [] has joined #linode
11:10|-|mauro_ [] has joined #linode
11:10|-|sracatus [] has quit [Ping timeout: 480 seconds]
11:14|-|mauro_ [] has quit []
11:37|-|schultmc [] has joined #linode
11:38|-|FireSlash [~FireSlash@] has quit [Read error: Connection reset by peer]
11:53|-|jekil [] has joined #linode
12:18|-|afv-13 [] has quit [Ping timeout: 480 seconds]
13:02|-|SpaceHobo [~spacehobo@] has quit [Ping timeout: 480 seconds]
13:09|-|afv-13 [] has joined #linode
13:37|-|spr_ [] has joined #linode
13:42|-|AnMaster [~AnMaster@] has quit [Quit: leaving]
13:46|-|gpd [~gpd@] has joined #linode
13:47<gpd>caker: $20 256Mb 10Gb 40Gb Xen ...
13:47<gpd>How is Linode Xen been going of late?
13:48<@caker>They say Xen is beta quality now ...
13:48<@caker>(this is as of Xen 3.0.3)
13:48<gpd>who is they? Xen people?
13:48<@caker>as far as competition, just wait a few months :)
13:48<gpd>do you still only have host 56 right now?
13:49<gpd>and the beta is closed...
13:49<@caker>You're welcome to move over to it
13:49<gpd>which xen is running on there?
13:49<@caker>it'll be 3.0.3 tonight or tomorrow
13:49<@caker>xend is locked up on it anyway, so it needs a reboot
13:50<gpd>any solution to the i/o thrasher problem?
13:50<@caker>not that I've implemented.. so far we've just been using cfq's ionice
13:50<gpd>i heard a rumour that raid1 helps
13:51<gpd>I have a friend running a map server on my linode and the i/o and apache CPU is starting to be heavy
13:51<gpd>i wonder if Xen will help with the i/o or not
13:53<gpd>caker: any hints on thw will be happening in 'a few months' ?
13:54<kvandivo>Hardware: Quad core, 64-bit machines (8+ ghz) man.. those are some fast machines
13:54<@caker>gpd: upgrades
13:55<gpd>nice -- another caker present :)
13:55<gpd>well - hopefully it will come soon as those slicehost prices are SHOCKING
13:59<gpd>on a different note: do people use postfix-gld for spam reduction?
13:59<gpd>I am approaching 400 per day on my account alone
14:00<gpd>which is probably contributing to recent linode load issue
14:01<kvandivo> had 3.23 million emails on a random day recently.. 77.2% were stopped at the boundary to the network and flagged as spam..
14:01<npmr>gpd, i use postgrey, it helps a lot
14:02<gpd>i tried postgrey but my REAL messages were delayed badly
14:02<gpd>thunderbird was good enough until recently - but now that is failing
14:02<npmr>how would postfix-gld get around that?
14:02<gpd>not sure -- but it has a whitelist
14:03<npmr>so does postgrey
14:03<gpd>so they are essentially the same thing?
14:03<gpd>ok -- so i should try postgrey again and this time get the whitelist
14:03<npmr>of course there are different implementation details
14:04<npmr>perl vs. c
14:04<npmr>bdb vs. mysql
14:04<npmr>probably others
14:04<fo0bar>caker: what baytechs do you have again?
14:04<npmr>gpd, senders should each only be delayed the first time they try to send a message
14:04[~]fo0bar is looking at these for work:
14:05<npmr>after the grace period is over, they should be automatically whitelisted
14:05<fo0bar> <-- photo
14:05<gpd>npmr: ok - sounds good - i obviously panicked last time
14:05<fo0bar>ebay has them for $250, but I'm need to find what they are retail
14:05<npmr>SupaZubon tells me it doesn't seem to matter whether the grace period is 30 or 300 seconds
14:05<npmr>i still use the default 300 seconds
14:05<@caker>fo0bar: I've got RPC4s, same thing, half the plugs
14:05<@caker>along with some zero-U ones (broomsticks)
14:06<fo0bar>caker: and they were all from ebay, correct?
14:06<@caker>fo0bar: those were yes (I did actually purchase DC72/DC74 modules from baytech a long time ago)
14:07<fo0bar>I'm probably just going to get straight IP units and put them in a dedicated private vlan
14:07|-|FireSlash [~FireSlash@] has joined #linode
14:07<fo0bar>assuming I can get funding for this :P
14:07<@caker>heh, I'll never trust any Baytech unit with networking again
14:07<@caker>Pretty sure for new deployments I'll still stick with Baytech RPCs, but I've switched to's console unit (rocks)
14:08|-|afv-13 [] has quit [Quit: The conversation has been moved to the Trash.]
14:08|-|spr_ [] has quit [Quit: "Ah, the bounce has gone from his bungie"]
14:08<gpd>npmr: can you tweak postfix|courier to use less RAM?
14:08<npmr>not so much with postfix (is that really an issue?), dunno squat about courier
14:08<fo0bar>caker: yeah, I don't like baytech too much because of their past problems, but the 2 big players in the "RPC units that don't cost 10 grand" market are baytech and servertech
14:09<fo0bar>and servertech doesn't have anything in the form factor I need
14:09<fo0bar>(our cabinets won't be able to take broomsticks, so I need 1U/2U)
14:09<npmr>gpd, most of the i/o usage in your mail processing will occur during message delivery
14:09<gpd>well i have 5 couriertls and 4 authdaemond.pla threads + a bunch of impads, pickup, tlsmgr, qmgr -etc... all using about 1.5M each
14:09<npmr>gpd, the sooner in the process you can reject spam, the less i/o it will incur
14:10<fo0bar>ironically, servertech's main office is about a half-mile down the road
14:10<npmr>yeah, ok 1.5M each for the *postfix* parts is low
14:10<npmr>i don't know what's typical for courier or how to tweak it
14:10<npmr>i don't use it
14:11<fo0bar>caker: thanks. I'm waiting back for a sales call to get "official" pricing on the RPC9
14:11<gpd>ok - no problem - i'll play with postgrey again - thanks
14:11<npmr>also, your threaded stuff are all sharing the same chunk of memory
14:12<npmr>if you've got 4 authdaemon.pla threads each showing up as using 1.5M, that's not per-thread
14:12<npmr>that's for the whole process
14:12<gpd>yeah -- unfortunately apache2 and mysql are still the hogs all this other stuff is trivial
14:12<npmr>apache2 prefork or worker?
14:12<gpd>ssh -> dropbear i just did - which might help a little
14:12<gpd>apache2 prefork - with all settings minimized
14:13<gpd>but php requires 16M for some php stuff :(
14:13<npmr>are you using any modules that would interfere with worker?
14:13<gpd>and I have 10 virtual hosts
14:13<npmr>ah, php
14:13<gpd>modules gone
14:13<gpd>not much more i can do i think
14:13|-|jimcooncat [] has joined #linode
14:14<gpd> <-- this is the mapserver
14:14<AnMaster_>caker, still null-routed?
14:14<npmr>mysql is always a chore
14:14<gpd>unless i can cache all of the 100M map I am screwed
14:14<@caker>AnMaster_: shouldn't be
14:14<AnMaster_>caker, hm, can't get in
14:14<npmr>if it's not allocating a bunch of ram and caching stuff, it's rereading rows from the on-disk tables
14:14<npmr>i/o either way
14:14<AnMaster_>ah now it works
14:15<npmr>your best option there is to make sure your queries are all as optimal as possible
14:15<npmr>indexing is your friend
14:16<npmr>ok, food
14:16<gpd>k - me too -- thanks again
14:19<gpd>Oct 31 10:44:52 www kernel: line_write_room: tty0: no room left in buffer ???
14:27|-|schultmc [] has quit [Ping timeout: 480 seconds]
14:30|-|SpaceHobo [] has joined #linode
14:43|-|FireSlash [~FireSlash@] has quit [Read error: Connection reset by peer]
14:45|-|schultmc [] has joined #linode
14:48|-|jimcooncat [] has quit [Quit: Leaving]
14:56|-|internat [] has joined #linode
15:12|-|internat [] has quit [Quit: This computer has gone to sleep]
15:26<gpd>npmr: where do you set the postgrey timeout?
15:27<SpaceHobo>gpd: I do it in /etc/default/postgrey
15:27<SpaceHobo>I set it to 1 or 5
15:27<SpaceHobo>because the escalating timeout setups create more problems than they solve
15:37<erikh>I'm pretty sure authdaemon uses a forked server model, not a threaded one
15:38<gpd>ok: --delay=30 for now , what about --max-age ? default ok?
15:38|-|spr_ [] has joined #linode
15:39<gpd>is there an easy way to monitor what messages don't retry? -- other than hand parsing /var/log/mail.log ?
15:40<erikh>eh, it will likely not be worth it.
15:40<gpd>so how else can i check that everything is working and good email is getting through?
15:40<erikh>greylisting more or less works on the principle that the people sending you mail are using something more sophisticated than telnet
15:41<erikh>gpd: .... you do know how greylisting works, right?
15:42<gpd>maybe: message gets rejected -- try again later -- spam doesn't bother
15:43<erikh>so your "good" email is going to get through as long as the people sending it have relatively clueful administrators, because temporary failure responses should always generate a re-attempt by the sending mail server
15:43<erikh>it really is that simple
15:43<gpd>but your sentence has 'relatively clueful administrators'... ;)
15:43<gpd>hence why a sanity check would be nice
15:44<gpd>the *actual* delay would be interesting
15:47<erikh>um, well, if the administrator on the other side is running a mail server that's less than 10 years behind on the ESMTP spec, you can safely bet that they're "clueful"
15:47<erikh>and the re-delivery delay does depend significantly on the other server
15:48<gpd>ok - cool
15:49<erikh>it's actually pretty hard to lose an outgoing mail queue these days
15:49<gpd>so how long before the spammers get around the greylist?
15:50<gpd>if they just look for a delay message they have a valid address?
15:51<erikh>why bother going to that extent?
15:51<erikh>they just have to send the mail twice
15:51<erikh>but they won't.
15:52<erikh>the delay message is applied before local delivery
15:52<erikh>IOW, they get temporary failure before valid addresses are resolved
15:52<erikh>of course, if you have VRFY turned on, well...
15:55<gpd>so the delayed messages should appear in mail.log - yes?
15:55<erikh>VRFY is a SMTP command that lets you check if a mailbox exists on the server
15:56<erikh>you probably don't have to worry about it, but it wouldn't hurt to put this in your
15:56<erikh>disable_vrfy_command = yes
15:57<gpd>ok - done - thanks :D
15:57<erikh>yeah, I know on sendmail at least you have to go through great pain (and turn on the "don't blame sendmail" macros) to even turn it on.
15:57<gpd>I keep getting a lot of these messages in mail.log too:
15:57<gpd> postfix/anvil[25086]: statistics: max connection rate 1/60s for (smtp:72.177.54
15:58<gpd>any reason to worry?
15:58<erikh>postfix periodically dumps stats. I would love to be more help with that, but I'm still relatively new to using postfix.
15:59<gpd>ok - final question: should I look for a whitelist for postgrey - or just rely on the auto-generated one?
16:17<erikh>honestly, I implemented postgrey yesterday on many recommendations to axe my current SA setup
16:17<erikh>between yesterday @ 10am and when I implemented it at 4pm, I got 62 pieces of spam
16:17<erikh>I have recieved 2 items that have landed in my junk box since
16:18<erikh>no major delays on delivery that I know of
16:18<erikh>SA is still setup, but that's because I'm swapping it out with DSPAM
16:20<gpd>ok - well it looks like it is working now - so fingers crossed: thanks erikh :)
16:20|-|Kurt [] has joined #linode
16:25|-|donnchao1aoimh [] has joined #linode
16:26|-|donnchaocaoimh [~donncha@] has quit [Ping timeout: 480 seconds]
16:31<gpd>well a double whammy already: spam from to an address that is similar to a yahoo address I use - but only 4 letters the same!
16:32<gpd>obviously it is relayed from yahoo -- so the greylist doesn't work :(
16:32<gpd>a yahoo is quite happy to wait 30 seconds
16:33<gpd>i suppose i'll just have to quit forwarding from the yahoo account
16:57|-|andrew_j_w [] has quit [Remote host closed the connection]
16:59|-|linville [] has quit [Quit: Leaving]
17:01|-|Newsome [] has joined #linode
17:15<gpd>arse - i am also forwarding from my .edu account - :(
17:19<gpd>looks like you can use postgreyreport to analyze the shizzle btw :)
17:22|-|jekil [] has quit [Quit: Leaving]
17:25<gpd> cat /var/log/mail.log | postgreyreport |less -S
17:25<fo0bar>caker: $800 retail, $650 new through reseller, $250 off the back of the truck ;)
17:25<fo0bar>(re: baytechs)
17:27<gpd> <-- this thing is *sweet*, my productivity levels are ^^
17:43|-|schultmc [] has quit [Ping timeout: 480 seconds]
17:46|-|schultmc [] has joined #linode
17:49|-|jekil [] has joined #linode
17:49<@caker>fo0bar: nice .. you guys have 20A circuits
17:49|-|spr_ [] has quit [Quit: "Ah, the bounce has gone from his bungie"]
17:59|-|Netsplit <-> quits: Dreamr_3, jekil, SpaceHobo, fo0bar, lucca
18:00|-|guinea-pog [] has joined #linode
18:01[~]guinea-pog bitchslaps HE
18:01<guinea-pog>caker: bitchslap them for me? thanks
18:02<guinea-pog>i've gotta run
18:03<Eman>i seriously hope its not that damn ddos again
18:03<Eman>its about the same time as last nights
18:04[~]caker calls
18:05|-|Redgore [~Redgore@] has quit [Ping timeout: 480 seconds]
18:05|-|taupehat [] has quit [Ping timeout: 480 seconds]
18:06|-|guinea-pig [] has quit [Ping timeout: 480 seconds]
18:06|-|fapestniegd [] has joined #linode
18:06<fapestniegd> again?
18:07<@caker>they're looking into it
18:07|-|sweh [] has joined #linode
18:07|-|DigitalM [] has joined #linode
18:07<fapestniegd>thanks caker
18:07|-|Dreamr_3 [] has joined #linode
18:07<guinea-pog>heh, here comes the cavalry
18:07<DigitalM>is host11 dead?
18:09<fapestniegd>DigitalIM traceroutes are stopping in's network (I'm on host18) they are looking into it.
18:09<DigitalM>argh... again!
18:09<DigitalM>thanks fapestniegd
18:10|-|claire{H-0}|away [] has joined #linode
18:10<Eman>heya claire
18:11<Eman> is having issues again
18:11<linbot>New news from forums: Network Issue in System and Network Status <>
18:11|-|fo0bar [] has joined #linode
18:11|-|hirstpf1 [] has joined #linode
18:12<hirstpf1>hi folks anyone on host23 having prolems right now?
18:12|-|zibeli [~zibeli@] has joined #linode
18:12<DigitalM> is having some issues hirstpf1
18:12|-|wirehead [] has joined #linode
18:12<Newsome>bah...HE issues
18:13<wirehead>bah...HE issues
18:13<hirstpf1>oh so I'm not the only one then!
18:13|-|zibeli [~zibeli@] has quit []
18:14|-|fapestniegd [] has quit [Quit: CGI:IRC (Ping timeout)]
18:14|-|Kurt [] has quit [Quit: Welfare recipients are the new feudal lords--they are the unproductive forcibly taking from the productive]
18:16|-|timothy [] has joined #linode
18:16<timothy>anyone here?
18:17<wirehead>are you here for the HE issues?
18:17<@mikegrb>mmm cake
18:17<wirehead>or just the cake?
18:17<sweh>We are; HE isn't
18:17<timothy>no, I'm having issues w/ my linode
18:18|-|BrianHV [] has joined #linode
18:18<Eman>timothy: unresponsive?
18:18<timothy>yesterday and today
18:18<Eman>most of us are having the same problems, its being investigated
18:18<sweh>It's a network issue at HE
18:19<sweh>Always worth checking and the forums in cases like this; many people notice and it gets mentioned :-)
18:20<@caker>so far, looks like HE's uplink switch
18:20<@caker>he's looking at it now
18:20|-|kathy|afk [] has joined #linode
18:21|-|guinea-pig [] has joined #linode
18:21<hirstpf1>must be those spammers on HE again ;)
18:21|-|darkness` [] has joined #linode
18:21|-|Kurt [] has joined #linode
18:21<timothy>right on
18:22<timothy>are they sucking the bandwidth? or damaging the CPUs somehow?
18:23<wirehead>naw, just impregnating your daughter and making your organ smaller.
18:23<hirstpf1>dunno but I have my suspicions because HE has lots of mentions for its IP ranges in one of the spam warning blacklists so I suspect they get targeted every so often
18:23<timothy>wirehead: lol
18:23|-|host60 [] has joined #linode
18:24|-|kioma [] has joined #linode
18:24<host60>any news on the host60 HE network issue?
18:24<timothy>right on
18:24<hirstpf1>being investigated
18:24<timothy>thanks for the info guys
18:25|-|sweh [] has left #linode []
18:25|-|taupehat [] has joined #linode
18:25<wirehead>oh well.
18:25<wirehead>I'll just wait for it to get back up. :/
18:25|-|wirehead [] has quit [Quit: CGI:IRC]
18:25|-|Redgore [~Redgore@] has joined #linode
18:25|-|timothy [] has quit [Quit: timothy]
18:26<hirstpf1>HE seems to have more problems than average with network
18:26<host60>what sort of problems are they having? Router meltdown or is their data centre on fire?
18:26|-|JDM [] has joined #linode
18:27<@caker>looks like a switchport died on their end
18:27<hirstpf1>ouch hope they carry inventory!
18:27<darkness`>that sounds like an extremely important switch port.
18:27<@caker>yeah, it was ours :)
18:28|-|SpaceHobo [] has joined #linode
18:28<hirstpf1>caker they giving an ETA yet?
18:28<hirstpf1>round of applause!
18:30<hirstpf1>yup working
18:31|-|GhostXz [] has quit [Ping timeout: 480 seconds]
18:31[~]gpd pressures caker to bring forward upgrades ;)
18:32|-|darkness` [] has quit [Quit: Leaving]
18:32|-|claire{H-0}|away [] has left #linode []
18:32|-|guinea-pog [] has quit [Quit: bye...]
18:37|-|kioma [] has quit [Quit: CGI:IRC (Ping timeout)]
18:37|-|host60 [] has quit [Read error: Connection reset by peer]
18:39|-|BrianHV [] has quit [Quit: CGI:IRC (Ping timeout)]
18:40<JDM>caker: more strange packets, but not dhcp queries "Nov 1 00:16:32 node0 kernel: IN=eth0 OUT= MAC=fe:fd:40:3e:be:5f:00:30:48:28:68:ec:08:00 SRC= DST= LEN=67 TOS=0x00 PREC=0x00 TTL=64 ID=18014 DF PROTO=UDP SPT=53 DPT=1320 LEN=47". they're not dns (port 53?) requests either because i can make requests via nslookup without that happening
18:41|-|kathy|afk [] has quit [Quit: it's probably windows update...]
18:45|-|lucca [] has joined #linode
18:45<lucca>well that was messy
18:58|-|JDM [] has quit [Quit: Konversation terminated!]
18:58<@caker>JDM: those are dns requets coming from (aka
19:21|-|DigitalM [] has quit [Quit: yay for me]
19:23<gpd>anyone use dropbear? I can't get publickey authorization to work
19:24<gpd>bah -- mv authorized_keys2 authorized_keys -- done
19:31|-|AnMaster1 [~AnMaster@] has joined #linode
19:31|-|AnMaster_ [~AnMaster@] has quit [Read error: Connection reset by peer]
19:41|-|Kurt [] has quit [Ping timeout: 480 seconds]
19:57|-|jekil [] has joined #linode
20:13<erikh>heh, I had no idea dropbear was that sophisticated
20:13[~]erikh will have to fiddle with his openwrt config now
20:14<erikh>caker: heh, I imagine tcpdump doesn't work so hot under UML, does it?
20:16<erikh>well shit, how about that
20:16<erikh>UML ftw
20:18<fo0bar>caker: yes, we have 20A circuits
20:18<fo0bar>however, I have no clue how much power we're using on each circuit
20:19<fo0bar>each circuit has a maximum of 12 devices on it, most are supermicro P4HT machines
20:19<fo0bar>so I'm pretty sure I could put a few more on each circuit, but it would be dangerous to try without knowing circuit draw
20:20<fo0bar>oh, and remote reboot would be nice too
20:31|-|hirstpf1 [] has quit [Quit: CGI:IRC]
20:32<@caker>erikh: tcpdump should work fine
20:33<erikh>caker: yeah, I was kind of surprised I guess
20:33<erikh>holy crap the trick or treaters are out in force this year
20:34<erikh>my wife and I wonder what spawned the plethora of ninja costumes
20:34<erikh>I mean, other than, well, it's ninjas.
20:37<erikh>ok, that kid was in a fencing costume
20:40<erikh>wow, can't even sit down before they knock on the door
20:55<warewolf>thou shalt not eggebuh.
21:00<warewolf>thou shalt not covet thy eggebuhs eggebuh.
21:00<warewolf>thou shalt not commit eggebuhry.
21:02<gpd>well -- postgrey 97 blocked - 14 slid through [so far]
21:13<npmr>heh.... authorized_keys2
21:13<npmr>that is so 2002
21:15|-|synchk [] has quit [Quit: Bai, fuckers.]
21:16|-|tasaro [] has joined #linode
21:24|-|tasaro [] has quit [Quit: Snak 5.1.5 IRC For Macintosh -]
21:45|-|FireSlash [~FireSlash@] has joined #linode
21:47|-|jekil [] has quit [Ping timeout: 480 seconds]
21:51|-|AfterDeath [] has quit [Ping timeout: 480 seconds]
21:52|-|Newsome [] has quit [Quit: Linux: Now with employee pricing!]
22:09|-|Dreamr3 [] has joined #linode
22:14|-|Dreamr_3 [] has quit [Ping timeout: 480 seconds]
22:15|-|AfterDeath [] has joined #linode
22:15<linbot>New news from forums: Date/Time on All pages in Feature Request/Bug Report <>
22:19<gpd>npmr: do you use denyhosts ?
22:21<gpd>i suppose if you just switch off all password access this is pointless
22:21<gpd>but the synchronization mode sounds interesting
22:23|-|FireSlash [~FireSlash@] has quit [Read error: Connection reset by peer]
22:27|-|schultmc [] has quit [Ping timeout: 480 seconds]
22:36[~]gpd mutters to self about slicehost
22:36<erikh>gpd: I just remove password authentication
22:57<tierra>caker: does serve as slave dns then (and that's automatically setup with domain additions)?
22:58<tierra>also, are those on separate hosts then?
22:58<tierra>(ns1 and ns2 I mean)
22:59|-|VS_ChanLog [] has left #linode [Rotating Logs]
22:59|-|VS_ChanLog [] has joined #linode
22:59<gpd>tierra: that is certainly what caker was talking about a while ago -- ie. ns1 and ns2 in different datacenters
22:59<tierra>gpd: thanks, haven't kept up on backlog
23:04<@caker>tierra: yes, each nameserver is in a differnet datacenters -- ns2 isn't online yet (but will be tomorrow)
23:06<gpd>caker: can you give us more details on these upgrades - I am intrigued
23:06<@caker>can't say just yet, sorry
23:06<tierra>caker: thanks, I'm switching my domains over right now, definately be happy to shutdown bind for the extra ram
23:06<@caker>tierra: cool .. it Works For Me
23:06<@caker>lemme know if you find something odd
23:07<tierra>will do
23:07<gpd> bind 20 0 30040 1436 992 S 0.0 1.0 0:16.16 named
23:07<gpd>doesn't look to be a huge saving :(
23:08<gpd>but will be very useful for some sites i have
23:09<tierra> bind 23 0 30560 3316 2152 S 0.0 1.7 0:04.09 named
23:09<tierra>looking at about the same
23:10<tierra>I swear I've seen it fork a lot more processes before and take up a lot more at times
23:11<tierra>either way, it's one less service I have to worry about
23:12<gpd>zactly :)
23:12<tierra>the DNS manager is looking nice btw, I like the UI and options
23:13<tierra>also, import from AXFR is a nice touch... very handy
23:13<@caker>it gets a little cumbersome, say, trying to delete a lot of A records from a domain
23:13<@caker>yeah, tricky :)
23:15<tierra>I'm not expecting to see all sorts of bulk edit options like GoDaddy's domain control center ;)
23:16<tierra>and I'm sure you probably don't want to encourage users signing up for a small base linode plan to start mass DNS hosting, heh
23:16<@caker>whatever :)
23:18<@caker>So, the interface makes fairly good sense if you've got DNS experience/used one of these before .. but it's going to be confusing to newbies
23:18<Eman>just label it "advanced users only"
23:18<@caker>needs some doco and some tooltips
23:18<@caker>A big friendly cube that comes out and suggests stuff
23:19<tierra>and his friendly assistant, Sphere
23:19<warewolf>"Hi, I see you're trying to add as a domain. While this actually will do absolutely nothing, you should know better. I've deleted the zone from your profile for you automatically."
23:20<Eman>It looks like you're writing a letter...
23:21<@caker>tierra: notice the domain modes? They're tucked away under the Domain view page -> Edit
23:21<@caker>Edit mode goes on serving the old zone file while you make edits, until you set it back to active
23:22<@caker>If a zone is auto-generated and is found to have errors, it changes to "Has Errors" and continues to serve the old domain, as well...
23:22<tierra>oh, nice, I was wondering what that was
23:23<@caker>s/old domain/old zone/
23:24<gpd>caker: so what's been happening in the last few months?
23:24<@caker>gpd: in what regard?
23:25<gpd>general linode news / caker ... ? I
23:25<gpd>'ve been afk
23:25<@caker>hehe .. not much, I guess .. we're still moving forward
23:25<gpd>i heard somethinga about a motorcycle incident...
23:25<@caker>ahh, yes
23:25<@caker>there's a URL, one sec :)
23:26<gpd>but all ok now I hope...
23:26<gpd>that was a while ago
23:26<@caker>yeah, tore my rotator, still not 100% movement in my shoulder
23:26<warewolf>my dad did that years ago
23:27<gpd>nasty -- straight back on another bike I bet though
23:27<@caker>um, what else .. I started running a few days ago. My plan is to get running part of my lifestyle ...
23:27<gpd>sweet -- i started running this summer too
23:27<gpd>surprisingly enjoyable --
23:28<@caker>gpd: heh, yes: :)
23:28<gpd>well compared to the 'end of the world' i assumed it would be
23:28<@caker>right, we got off lucky... that's about as "good" as getting hit by a car on a motorcycle gets
23:29<gpd>nice bike -- I see that it is red [knowledge == high]
23:29<gpd>all insurance paid i hope
23:29<@caker>what's new with yourself?
23:29<@caker>they did, they were good about it since it was her fault, etc
23:29<gpd>not much - been trying to finish up in LA and move back to UK soon
23:30<gpd>got engaged... hawaii... etc
23:30<gpd>thanks :)
23:31<gpd>where is the mikegrb these days?
23:34[~]gpd checks logs and finds mikegrb alive and well
23:35<tierra>Wow, I hadn't heard anything about the motorcycle incident til now, good to hear everything turned out ok. Looks like there could have been some serious injuries as far as your foot and hand excluding other factors.
---Logclosed Wed Nov 01 00:00:27 2006