tabthorpe: awesome, tnx! toddf: the newer libvirt _still_ does not support reboot. :( toddf: i will have to do even a further upgrade, but first i gotta get the initial upgrade going. baby steps :) ryk: hmm... man, why didn't i think about that. a chroot'd home may very well work. but wait, would that still preserve permissions with rsync? you need to be logged in with UID 0 for rsync to be able to set perms. Hmmm... i guess you could make the user UID 0 in /home/foo/etc/passwd bryc3c: toddf : IIRC, my KB does show cores at $2 per core the reason i think i hesitated from putting it in the power ups is b/c SMP isn't supported on all guests (like, OpenBSD barfs). maybe i need to get over that. Poor OpenBSD. to be honest, I don't remember what I meant by "Dedicated CPU resources", i wrote that like 3.5 years ago... ;) I probably was contrasting KVM/QEMU to something like OpenVZ, which doesn't really dedicate resources mercutio: "todd: i wonder if you're on a newer cpu than me" <-- negative. All 26 servers have the exact same specs Now, the *27th* server is the 2nd gen build that several of you have requested to beta I have had people report that multiple cores did in fact increase their app performance (e.g. Minecraft server) How many VMs on average run on each server? Just out of curiosity. it seems a lot easier to bottleneck a disk than a cpu; that is indeed where we see the bulk of load CaZe: it really depends on the server and nature of VMs. but it isn't hard to get 20 on a box. with a ton of cores and RAM, i bet you could do 60 VMs running screen+irssi pretty easily ;) but i've never tried Oh, that would be me right now. lol If I want to change to six month billing, do I just create a support ticket? CaZe: yup which leads me to beg the question: if i want to switch to no billing (re: free vps) do i just create a support ticket? lol phlux: i think that is equivalent to cancelation ;) up_the_irons: cpu is l5420 or something isn't it? up: you should be able to do 200+ vm's with screen/irssi screen/irssi 256mb 200 on a 64gb box like 8x8gb you dont need 256m ram for irrsi :) a slim kernel and tiny linux shoudl work with 56m or something yeh until you run apt-get up_the_irons: i think you did mention the rsync difficulty last time we talked about that. call them up and ask how they get around that :) you guys should totally send the vps ssh fingerprints in the welcome email hai ryk for i in $( ls /etc/ssh/ssh_host_* ); do ssh-keygen -l -f $i; done somelikethat and sign the same email, since if you're gonna be paranoid... ;) heh so - a new vps i just got seems to have taken out an ip of have in one of my blocks qbit: please try English again i had two vps's one with extra 5 ips and i just got a new vps ( technically 4 because the 5th was used by said vps ) but the new vps used teh 4th ip from that block reducing my number of ips from 4 to 3 as is policy iirc sadpanda mk but imma leave it up to up_the_irons to answer for reals since I haven't found what I'm looking for, yet closest I found was http://support.arpnetworks.com/kb/vps/are-private-ip-addresses-available-for-vps-to-vps-communication qbit: does you show an additional block under IP_BLOCK in your portal? Maybe you were given another address, but it automatically setup the VPS with a different one (first available, from your /29) in which case, just change the IP settings and be happy i do have IP_BLOCK but the new vps used the nexta avail from said block qbit: and IP_BLOCK doesn't listed any new addresses? use your english! damnit 10:10:32 ::: #arpnetworks You're not a channel operator no :P it lists the range it should - but the new vps took next avail ha pwnt *doesn't list any new or additional IP addresses? no i had 4 - now i have 3 Well then.. either wait for up_the_irons to respond, or email support@... or maybe someone else here with more knowledge can help hehehe devio.us qbit: eh? what do you mean... we have admins@devio.us and I see every one of those emails. update support it's still 4.7 yeah that's not the same problem hence the quotes and bitch all you want... apparently it's not going to do anything yep Did you guys ever get that premium server going? nope well "going" CaZe: the servers are up and running... bryc3c has logins but no one else does But we don't have the new admin stuffs written and therefore cannot add accounts or anything Also this hardly seems like the right place for a #devious discussion yeah - your comment about the service reminded me i was gonna cancel :D heh :) ryk: lol, or just order a vps ;) qbit: that's not a bad idea. email feedback@arpnetworks.com and i'll put it in the feature requests queue :) qbit: bryc3c is right; IPs for any / all services come from your account's IP block(s), they are not assigned to individual VPS' cool up_the_irons: but does ordering an additional VPS not include an additional /30? brycec: no, it does not. first service of anything you order (vps, colo, dedi, etc...) includes a /30, then after that, IPs must be purchased (usually $1 per IP) a /30 per VPS would waste a lot of IP space heh, gotit. Exactly what I was trying to confirm (for qbit) :) :) yay thanks for the clarification np up_the_irons: Same question on IPv6 - /64 shared across all VPS? lol if you need more than a /64 mind blown brycec: yup must have http://www.welovefine.com/3471-what-happens-on-the-holodeck.html ha that's awesome haha a certain busty friend of mine has a similar shirt - LCARS design with the same saying. lol the "expendable" one is awesome too thanks up_the_irons for handling the upgrade (if you hadn't matched IRC nick to email yet...) so... let it be known .. that even though openbsd introduced a fix for some broken soekris mpbios that should help arpnetworks kvms as a side effect, the real effect is that you can install and boot a standard kernel on arpnetworks kvms yet it gets watchdog timeouts on em0 and network interfaces and then freezes. probably chews a bit of cpu also. the fix? the same old workaround: 'disable mpbios' ;-( with the 'disable mpbios' even 'virtio' networking works fine oh, they have virtio networking too? as of post 5.2 disk support too though bios in the current production kvms at arpnetworks does not support hooks to boot openbsd off virtio disk (apparently freebsd also) brycec: np! toddf: doh, figured that wouldn't make it into 5.2 brycec: yeah i made the nick connection when I'm done decompressing from moving my office and now colo-move in progress, I'll focus on getting it into the installation media also .. if someone else doesn't beat me to it ;-) toddf: it'll be interesting if the beta VM (newer kvm/qemu/bios) will fix the mpbios issue now that I have working net and disk (after futzing with tests up_the_iron was gracious enough to let me do) its full steam ahead to move my colo to arpnetworks ;-) how much colo? I'm getting kinda low on space atm... cd $food up_the_irons: apparently its a buggy release of the bios, so it'll just depend on whether or not the bios is past that release fix or not up_the_irons: my 2nd vps that I was futzing with the virtio disk/net recently will replace my colo .. so no worries its just a matter of me having time to transfer services, renumber various things in dns after verifying things are up and working, you know, typical seamless migration strategies and all @up_the_iron+│ phlux: i think that is equivalent to cancelation ;) booo no freebies for you! oh todd you got network going with virtio! is there a workaround for the disk ting? i care about network more then disk :) they're already doing post 5.2 changes the changelog show that now, oops :) toddf: the newer kvm/qemu uses a whole new bios (SeaBIOS instead of Bochs) toddf: ah, to replace colo, good idea phlux: lol why oh why, Lucid, did you break the 'inet6 static' stanza in interface(8) ? *interfaces man, i was using kdm to log me into i3, and i found out it was using 100% of one of my cpus my load averages were in the 1.5 range with i3, 4 instances of urxvt, google-chrome, thunderbird, and keepassx. i knew something was wrong with that. disabled kdm and networkmanager, and i'm sitting at 0.12 now with all of that loaded free -m: -/+ buffers/cache: 604 7136 woot that's awesome up_the_irons: Have you seen novnc? It's a javascript/html5 vnc client. Seems like it would be a good fit for the portal. http://kanaka.github.com/noVNC/noVNC/vnc.html bryc3c: i think i have bryc3c: something like that will eventually make its way to the portal bryc3c: that looks pretty sweet actually.. i'd need that bridge thingy though wow, the screenshots show it used with an ipad now, that's, hot