[00:00] ouch [00:02] did you still run well before the herniated disc? I'm trying to get my 1.5 mile time down to 10 mins (right now it is 12:30) [00:02] Yep, I did pretty well [00:02] Our PT test is actually 1.5 mile [00:03] we had to do it in 12:50, which seems like a lot of time to me [00:03] I played baseball in high school, so when I joined the CG, 12:50 was easy [00:03] yeah, that seems like a lot [00:03] I did it in 11:00 [00:03] nice [00:03] the swimming part is a bitch [00:04] it used to be 150 yards, but we had to do 500 [00:04] hah [00:04] water and me don't mix [00:05] really cold water makes me unable to breath [00:05] well that's why the Coast Guard is here for you :D [00:05] LOL :) [00:05] yeah they kept the pool water in basic training at a steady 65 [00:05] it was freezing [00:06] i could do 65 i think; once i was in catalina water at like 50; well F*CK that [00:06] couldn't breath [00:06] yeah 50 is bad [00:07] the last thing you want is hypothermia [00:07] we have these "mustang" suits for that purpose [00:08] they float, keep us dry, and keep us warm using our body heat [00:08] that was the only fun part about basic.. learning how to use those [00:08] hmm..brb [00:09] ah cool [00:35] hey up_the_irons [00:36] hey RandalSchwartz [00:36] so I got all my IPs on both boxes working now [00:36] including my v6's [00:37] sweet [00:37] just realized that's 80 bits worth of personal addresses [00:37] more like 16 bits worth of subnets [00:38] indeed. [00:38] and with comcast starting to deliver v6 to homes... [00:38] in IPv6, "number of IPs" is actually not much of a topic [00:38] I can use those [00:38] b/c there are so many; instead, we focus on how many subnets we have [00:38] /64 is your standard single subnet [00:39] i'm glad to see comcast getting on the ball [00:39] wonder if they'll give you a /56 or /48 to your home [00:39] what is mzima.net? [00:40] one of my transits [00:40] ok [00:40] I notice my traceroutes seem to be different at different times. :) [00:41] just whatever's closest, right? [00:41] "closest", yeah [00:41] anyway, I've got the formula down for how to turn your info into freebsd rc.conf files [00:41] it'll tend to fluctuate, just the nature of the 'net [00:42] I wonder if you could have a wiki so I can share what I figure out [00:42] not essential... just a thought [00:42] mmm.. all i got in the knowledge base [00:42] random thoughts are like that. :P [00:42] a wiki is a good thought though [00:43] in fact, you could modify your template for the box page to show configs [00:43] so I could just copy-paste them [00:43] if you knew I was freebsd, that is [00:43] oh that's true [00:43] yeah, the portal knows [00:44] i have an old article: http://support.arpnetworks.com/faqs/main/how-to-configure-ipv6-on-freebsd [00:44] defaultrouter=174.136.96.177 [00:44] ipv4_addrs_em0=174.136.96.178/30 [00:44] gateway_enable=YES [00:44] ipv6_enable=YES [00:44] ipv6_gateway_enable=YES [00:44] oops [00:44] ipv6_defaultrouter=2607:f2f8:1640::1 [00:44] ipv6_ifconfig_em0=2607:f2f8:1640::/48 [00:44] like that [00:44] if you published that based on the numbers, I coudl just copy/paste it [00:45] you need that when you're building the box stock anyway [00:45] or how do you image it? [00:45] those are generated automatically upon provisioning [00:45] rc.conf, etc... [00:45] oh cool [00:45] maybe if I had just captured mine before wiping the disk... :) [00:46] haha [00:46] it gets a little more complicated when a customer has more than 1 vps [00:46] i'm looking to separate the ip block info from the vps service item [00:46] I've talked with a lot of people about the idea of offsite ZFS snapshot backup service [00:46] you really need to look into that [00:46] it'd a be service almost nobody else is offering [00:46] b/c multiple VMs per customer is becoming a pain [00:47] yeah, i like that idea [00:47] hey - how do you handle someone wanting to go from a 20GB to a 40GB drive? [00:47] do you mount a 40GB drive in addition for a bit? [00:47] no [00:48] lvextend, fdisk -BI, disklabel, growfs [00:48] in a NUTSHELL [00:48] does zfs handle that? [00:48] oh i have no idea how to do it on zfs [00:48] most upgrades are ufs2 [00:48] yeah, just wondering [00:48] the ones that are not, are ext3 [00:48] might need to look into that [00:49] not for me, but neil might eventually get rid of $90/month to amazon [00:49] and send it your way instead [00:49] ec2? [00:49] yeah [00:49] what kinda stuff do u get for $90 on ec2? [00:49] just your basic dual CPU virtual box [00:50] running linux, or..? [00:50] opensolaris or linux or windows [00:50] not freebsd [00:50] ah i c [00:50] funny they support opensolaris first before freebsd [00:50] I'm going to mirror his stuff to his box here [00:50] Sun did it [00:50] ah [00:50] Sun made sure they were a player on EC2 [00:51] haha [00:51] the fbsd people are trying to get their stuff together [00:51] but not yet [00:51] yeah [00:51] so neil is looking for a ZFS based solution? [00:51] I'm gonna mirror all of his website from EC2 to here [00:51] (off-site backup, that is) [00:51] and then fire it up [00:51] and show him it's just as good [00:51] so he can get rid of EC2 [00:51] sweet [00:51] but his 20G is likely going to be a bit tight after a while [00:51] he has some media files to serve [00:52] hence the question [00:52] ah [00:52] and you put on ZFS for him? [00:52] his box is already live with email and DNS [00:52] yeah [00:52] ok [00:52] interesting [00:52] if you could just add a second 20G to the provision, I can mount it and spread data across both [00:52] zpool works that way [00:53] not sure what that looks like outside th ebox [00:53] i mean, i think it should be straight forward... if ZFS supports growing itself, then it is not an issue; extending the backing device (e.g. "making the 'physical' hard drive bigger") is trivial [00:53] yeah, I'll look into that [00:53] might require a few zraid swaps [00:53] yeah, alternatively, i could easily give you a 2nd disk [00:53] 20G zraid with 40G will silver 20G on both [00:53] in fact that is the way i did upgrades before i was savvy with growfs and e2resize [00:54] then you "fail" the 20G [00:54] and replace it with a 40G [00:54] and it silvers both with the data [00:54] then you "fail" the second 40G [00:54] nice [00:54] so that's one way to grow a disk if you have a bit of time [00:54] the nice thing is you can do that "live" [00:54] if you were to google for your remote zfs backup solution, what terms would you search for? [00:55] yeah, the "liveness" of it is hot [00:55] traditional method requires downtime [00:55] zfs is all about the uptime :) [00:55] zfs offsite backup [00:55] and there's already someone claiming to be "first" [00:55] roger [00:55] damn! [00:55] haha [00:56] ... http://www.engadget.com/2008/07/29/dattos-z-series-worlds-first-on-off-site-backup-solution-to/ [00:56] two years ago! [00:56] smugmug uses ZFS [00:56] I should get you in contact with someone there [00:57] so you can talk to them about ZFS best practices, and maybe they can help you figure out how this is best done [00:57] anyway - I think there's a business in there [00:58] and if you can figure out how to solve the last piece of redundancy [00:58] yeah i think you're right [00:58] you'll be one of the few (if not the first) to have it all [00:58] http://developers.sun.com/solaris/articles/storage_utils.html [00:58] sun has something to say about it [00:58] you already have redundant disk, power, nets, etc. [00:58] you just need redundant geography [00:58] and I know many corps would pay nicely for htat [00:58] at multiple levels [00:59] data somewhere else [00:59] data somewhere else, ready to become hot after DNS [00:59] redundant geography isn't too hard; just gotta ship a router and some boxes somewhere [00:59] data somewhere else, ready to become hot on HA failover [00:59] and with ZFS snapshot shipping, minimum bytes to ship [01:00] the 2nd site would have different IP range, so it'd be a DNS fail-over solution; with some experimentation and cost, it may be possible to share IP space [01:00] well - as I said, you could price differnet solutions differently [01:00] right [01:00] for me, for example, if I just had a way to get my data again, even if it required provisioning a new box, I'd be happy [01:00] but for some, having a second spare box ready to fire up would be better [01:01] right [01:01] that is, for me, you wouldn't have to allocate enough CPU beyond receiving the ZFS data [01:01] but for the second tier, you have to allocate enough cpu/memory/power for a box to go live at any moment [01:02] that's why I think you could have a nice variety of products for every off-site need [01:02] at different price points [01:03] I'm suggesting this to you because it's clear to me you already understand and live and breathe redundancy. :) [01:04] Yes :) [01:04] I think about redundancy every day, it is a sickness really [01:04] ;) [01:05] well - you're the guy I want on my side then. :) [01:06] by the way, I landed a nice 3-month contract in LA starting 3/15 [01:06] so I'll be down your way quite a bit [01:06] do you know oversee.net ? [01:06] nice [01:06] no, i'm not familiar with them [01:07] basically, about 80% of the search-spam landing pages... [01:07] all done with a few dozen linux boxes to serve something like a million domains [01:08] nice [01:08] 44th floor of the building just northwest of the downtown library [01:08] some odd lawfirm name [01:08] diagonal from mccormick and schmicks [01:08] 4th and flower maybe [01:08] something like that [01:09] 515 S Flower [01:09] sounds familiar [01:09] so anyway, I'll be nearby soon [01:09] for probably halftime onsite, halftime off [01:10] I'd like to meet you some day. :) [01:10] sure, let's get some lunch :) [01:10] and you'll be real close to the data center, i can show you around [01:10] nice [01:11] the center that's also the "cop station" in southland :) [01:11] yup, exactly :) [01:12] ok - well, I think it's time for me to sleep... nice chatting with you again. [01:13] RandalSchwartz: nice chattin' ! g'night [02:43] *** [FBI] starts logging #arpnetworks at Sun Mar 07 02:43:00 2010 [02:43] *** [FBI] has joined #arpnetworks [02:58] *** ziyourenxiang has joined #arpnetworks [04:01] today when i ran freebsd-update on 7.2 is says that 7.2 is EOL soon... what do peope run? RELENG_7 or RELENG_8? [04:37] I use 8.0 [04:42] 8 on my vps, 7 on my own colo box (need to update it!) [05:02] up_the_irons: openbgpd was written and developed to address the deficencies of quagga (and its predecessor) [05:02] up_the_irons: you'll find that pps performance is indicative of the hardware/os not the routing software [05:03] up_the_irons: openbgpd specifically has good bgp performance, not taking minutes to recalculate routes if anything changes; it is a tangent but it is noteworthy openbgpd takes 32mb for a full routing table, cisco takes 64mb [05:04] up_the_irons: I've never seen what you describe on MegaRAID it must be an older system or something, always clearly raid0/raid1/raid5/raid10/etc on my systems: [05:05] eclipse$ sudo bioctl ami0 [05:05] Volume Status Size Device [05:05] ami0 0 Online 8986296320 sd0 RAID0 [05:05] 0 Online 9106882560 0:15.0 noencl [05:05] ami0 1 Online 440087347200 sd1 RAID5 [05:05] 0 Online 146811125760 1:3.0 ses0 [05:05] 1 Online 146811125760 1:8.0 ses0 [05:05] 2 Online 146811125760 1:4.0 ses0 [05:05] 3 Online 146811125760 1:5.0 ses0 [05:05] ami0 2 Hot spare 146815733248 1:2.0 ses0 [07:37] ziyourenxiang: 7.2 is EOL soon because 7.3 will be out [07:38] Jesus, I should just pf deny everything from apnic. [07:38] ok, so i'll be able to freebsd-update to that then [07:39] Yep. 7.3 just isn't out quite yet. [08:02] *** ziyourenxiang has quit IRC (Quit: ziyourenxiang) [13:04] hm [13:15] hm [14:17] *** visinin has joined #arpnetworks [14:52] *** visinin has quit IRC (Quit: leaving) [15:51] *** amdprophet has quit IRC (Quit: Leaving...) [16:12] anyone in the channel have sales experience? [16:13] I sell myself all the time [16:13] * RandalSchwartz puts on the red dress and high heels [16:14] what's your question? [16:21] I'm a geek with no inclination to sales. I started a software consulting company, and got some projects in 2004-2005 [16:21] since then, things have been slow [16:21] I'm trying to figure out how to sell my services. I have a website, and have good SEO rankings [16:22] figure out what you're good at, and where people are who might want that [16:22] then figure out what they read, look at, think about [16:22] and get into those [16:29] yeah, i have been relying on writing articles and other things to 'attract' people, but i'm thinking I'll have to start cold calling [16:33] old calling sucks [16:33] in general, easier to go hang out where they are [16:33] conferences, watering holes [17:20] shmget, im in the same boat as you [17:20] i dont code and shit but [17:20] i do do well for myself, i just am not a salesman, it sucks ass [17:20] brb lazy [17:28] once in awhile someone sees my resume and gets in touch, but not much [17:36] *** baklava has quit IRC (*.net *.split) [17:38] *** baklava has joined #arpnetworks [18:20] *** BarberRonny is now known as hiram [18:37] *** heavysixer has quit IRC (Quit: heavysixer) [20:01] *** heavysixer has joined #arpnetworks [20:01] *** ChanServ sets mode: +o heavysixer [22:44] *** visinin has joined #arpnetworks