1198716923 M * Bertl okay, off to bed now .. have a good one everyone! cya!
1198716933 N * Bertl Bertl_zZ
1198719258 J * rob-84x^ rob@submarine.ath.cx
1198722616 Q * nysis Quit: Αποχώρησε
1198726421 Q * yarihm Quit: Leaving
1198726746 Q * FireEgl Read error: Connection reset by peer
1198727637 J * FireEgl FireEgl@4.0.0.0.1.0.0.0.c.d.4.8.0.c.5.0.1.0.0.2.ip6.arpa
1198728448 Q * derjohn Ping timeout: 480 seconds
1198728484 J * derjohn ~derjohn@dslb-084-058-203-069.pools.arcor-ip.net
1198729043 J * hparker ~hparker@lonegroved20p183-d.cld.brightok.net
1198729711 Q * hparker Quit: g'nite
1198732518 Q * ard6 Ping timeout: 480 seconds
1198733593 Q * phedny Ping timeout: 480 seconds
1198740367 Q * AndrewLee Quit: leaving
1198740993 J * AndrewLee ~andrew@flat.iis.sinica.edu.tw
1198741709 J * ard6 ~ard@2002:d9c4:2909:1::1
1198742193 Q * ard6 Ping timeout: 480 seconds
1198744021 J * ard6 ~ard@2002:d9c4:2909:1::1
1198744957 N * DoberMann[ZZZzzz] DoberMann
1198745184 J * meandtheshell ~sa@85.127.102.217
1198745838 J * JonB ~NoSuchUse@0x535f65c3.kjnxx7.adsl-dhcp.tele.dk
1198746357 Q * Flinx Remote host closed the connection
1198747198 J * dna ~dna@p54BCC889.dip.t-dialin.net
1198749168 J * pmenier ~pme@LNeuilly-152-22-72-5.w193-251.abo.wanadoo.fr
1198749842 J * dna_ ~dna@p54BCC889.dip.t-dialin.net
1198749998 Q * dna Ping timeout: 480 seconds
1198750012 Q * dna_ Read error: Connection reset by peer
1198750043 J * dna_ ~dna@p54BCC889.dip.t-dialin.net
1198750090 J * phedny ~mark@ip56538143.direct-adsl.nl
1198750653 Q * dna_ Ping timeout: 480 seconds
1198750759 J * Pazzo ~ugelt@reserved-225136.rol.raiffeisen.net
1198750840 Q * JonB Quit: This computer has gone to sleep
1198751140 J * dna ~dna@p54BCD42F.dip.t-dialin.net
1198751278 J * JonB ~NoSuchUse@0x535f65c3.kjnxx7.adsl-dhcp.tele.dk
1198751657 J * flinx ~flinx@86.57.216.113
1198751766 Q * flinx 
1198754660 J * amax ~amax@r5ct9.net.upc.cz
1198754670 M * amax hi all
1198754677 M * amax one stupid question..
1198754714 M * amax what is easy and cheap way to read syslog of guests on host machine ?
1198754746 M * amax not read /vserver/ctx/var/log/messages .. but read his ctx's syslog
1198754905 M * amax eh..
1198754910 Q * amax 
1198755198 Q * JonB Quit: This computer has gone to sleep
1198756160 Q * Aiken Remote host closed the connection
1198756248 Q * dna Read error: Connection reset by peer
1198756273 J * dna ~dna@p54BCD42F.dip.t-dialin.net
1198756915 N * Bertl_zZ Bertl
1198756926 M * Bertl morning folks!
1198757055 J * pmjdebruijn ~pmjdebrui@jester.pcode.nl
1198757056 M * pmjdebruijn hi
1198757066 M * Bertl welcome pmjdebruijn!
1198757072 J * virtuoso ~s0t0na@ppp91-122-94-20.pppoe.avangard-dsl.ru
1198757120 M * pmjdebruijn Hi
1198757130 M * pmjdebruijn I'm trying to have one vserver use binfmt_support for Mono
1198757176 M * Bertl you have to configure it on the host then
1198757206 M * pmjdebruijn do those settings remain visible within the vserver then?
1198757217 M * pmjdebruijn Hmmm, I could have a go at it anyways, of course
1198757218 M * Bertl yes, they should, just be careful with the pathes
1198757227 M * pmjdebruijn Bertl: naturally
1198757236 M * pmjdebruijn anyway, thus far, I'm quite happy with vserver :)
1198757246 M * Bertl glad to hear that!
1198757268 M * pmjdebruijn the near-zero overhead is just a dream
1198757271 M * pmjdebruijn even on older hardware
1198757280 M * Bertl or embeded devices ...
1198757284 M * pmjdebruijn I have it running with Debian on a HP DL380 G2 (Pentium III 1400mhz)
1198757311 M * Wonka i'd be even happier if i had 2.6.24, a vserver patch with ipv6 for it, and an ipv6 /40 for the club...
1198757323 M * Wonka *g*
1198757340 M * Bertl Wonka: you should have something to test in a day or two
1198757348 J * JonB ~NoSuchUse@0x5739c847.roennqu1.broadband.tele.dk
1198757360 M * Wonka Bertl: no hardware suitable for testing here :/
1198757399 M * pmjdebruijn I'm hope to see .24 soon to
1198757405 M * Wonka only $there
1198757407 M * pmjdebruijn though I'm not that savy for IPv6 yet
1198757407 M * Bertl Wonka: well, then either take it without testing ... or find somebody to test :)
1198757423 M * Wonka anyway, i need some v6 addresses first
1198757435 M * JonB or get a virtual machine to test with
1198757448 M * Bertl IIRC, getting ipv6 addresses is quite simple
1198757477 Q * virtuoso_ Ping timeout: 480 seconds
1198757479 M * pmjdebruijn via a tunnel provider
1198757481 M * pmjdebruijn or your own
1198757500 M * pmjdebruijn for example the Dutch ISP XS4ALL provides tunnel+IPs for their customers
1198757573 M * Wonka my sixxs.net tunnel got deactivated some years ago because my mails bounced, because my computer club hat to be offline for a week...
1198757596 J * yarihm ~yarihm@84-75-119-160.dclient.hispeed.ch
1198757601 M * Wonka https://chaos.in-kiel.de/~w/sixxs_sucks.txt
1198757804 M * Hollow Wonka: i totally agree, they disabled my tunnels during my move to another city ...
1198757908 M * Wonka and they don't like freemailers like gmx
1198757938 M * Wonka so, what do they want? stable mail servers or non-freemailers?
1198758003 M * pmjdebruijn heh
1198758012 M * pmjdebruijn That's why GMail for Domains really owns
1198758059 M * Hollow it's just rediculous that everyone complains about the rare distribution of IPv6 subnets given the requirements ..
1198758067 A * ard6 suggests moving to 6to4
1198758071 M * ard6 it's *FAST*
1198758117 M * ard6 you can fix your own reversed resolving
1198758121 M * ard6 and stuff like that
1198758150 M * ard6 It's fast because on the 6to4 net the IPv4 is actually used as a "layer 2"
1198758206 M * ard6 And of course you get a /48 per public ipv4 adres you have :-)
1198758250 A * ard6 has a sixxs tunnel
1198758272 M * Wonka ard6: but my public v4 changes every 24h
1198758273 M * ard6 but the rtt to the tunnel is bad
1198758279 M * ard6 Wonka : ah, ok :-)
1198758285 M * ard6 Wonka : germany?
1198758290 M * Hollow heh
1198758290 M * Wonka ard6: and the one fix ip i have i don't want to publish more than necessary
1198758293 M * Wonka ard6: ack
1198758293 M * Hollow same here
1198758324 M * ard6 Well, I must tell you, it *is* possible to get it fixed :-). But it will cost you :-)
1198758383 M * ard6 we (the company I am employed at) have a german connection, and it is private and fixed...
1198758402 M * ard6 inetnum:        213.146.112.0 - 213.146.127.255
1198758402 M * ard6 netname:        KAMP-DSL-NET
1198758402 M * ard6 descr:          KAMP Netzwerkdienste GmbH
1198758402 M * ard6 descr:          DSL-Pool fixed IPs
1198758419 M * Wonka i could surely get a fixed ip too...
1198758433 M * Wonka am employed at an ISP myself
1198758444 M * ard6 ah :-)
1198758453 M * Wonka but "we" won't look into ipv6 very soon :/
1198758461 M * ard6 that's not a problem...
1198758462 A * pmjdebruijn too
1198758467 M * ard6 6to4 is the way out...
1198758473 M * pmjdebruijn we already provide tunnel for some customers
1198758482 M * pmjdebruijn though we're not fully IPv6 native yet
1198758487 M * ard6 that's something like 2002:<ipv4 adres>::/48
1198758490 M * pmjdebruijn doing that should be reasonably trivial
1198758508 M * Bertl pardon me, what is the advantage of ipv6 if you are still connected/tunneled via ipv4?
1198758519 A * ard6 is testing ipv6 with 6to4 : like www.ipv6.telegraaf.nl
1198758535 M * Bertl (besides the ability to use up the full bandwith less efficiently :)
1198758545 M * pmjdebruijn ard6: you're dutch :)
1198758550 M * ard6 :-)
1198758591 A * ard6 has a phonecall
1198758874 M * ard6  Anyway: we use 6to4 until we got our own /32 from RIPE :-)
1198758934 M * ard6 Bertl : Using 6to4 you are actually not tunneling anymore. All your services(vservers) will be available for the whole ipv6 world :-)
1198758941 M * ard6 and that world is growing
1198758994 M * ard6 (and for bandwith: we have 26Gb/s @ work ;-0 )
1198758998 M * Hollow well, it's all kind of useless as long as ISPs don't offer direct IPv6 connectivity
1198759039 M * ard6 they don't have to if you use 6to4. And then some point in time everbody is using ipv6, and then they have to..
1198759053 M * ard6 1337! Coffee :-)
1198759067 M * Hollow they don't have to, but they should asap, and nearly nobody does
1198759078 M * pmjdebruijn ard6: uh? you work at an ISP too?
1198759150 M * ard6 pmjdebruijn : I work @ the ISP for the telegraaf (contracted), and we have been becoming an ISP since 2000 or so :-)
1198759168 J * virtuoso_ ~s0t0na@ppp89-110-2-14.pppoe.avangarddsl.ru
1198759334 A * ard6 has the coffee :-)
1198759368 M * ard6 pmjdebruijn : We also host debian stuff. Actually we used to host security.debian.org, untill some major security update
1198759381 M * ard6 at that time it was too expensive to keep that.
1198759432 M * Bertl security update of whom?
1198759441 M * ard6 http://217.196.41.9/~ard/pics/2007-07-10/IMG_4245.html
1198759457 M * ard6 Bertl : Some major sarge security fix.
1198759505 M * ard6 everybody started downloading, and we didn't have enough peering, and the bandwidth price then was about 40 euro/Mbit
1198759530 M * ard6 That would have been 4000 euro a month extra
1198759547 M * ard6 Now we wouldn't really notice it :-)
1198759578 Q * virtuoso Ping timeout: 480 seconds
1198760181 M * pmjdebruijn ard6: yeah we like Debian too
1198760871 Q * JonB Quit: This computer has gone to sleep
1198761150 A * ard6 is going to put asterisk in a vserver
1198761161 M * ard6 with SYS_NICE :-)
1198761375 J * DLange ~DLange@212.16.242.195
1198761382 Q * dna Read error: Connection reset by peer
1198761407 J * dna ~dna@p54BCD42F.dip.t-dialin.net
1198762550 J * ema ~ema@rtfm.galliera.it
1198764080 J * Flinx ~chatzilla@86.57.216.113
1198764098 M * Flinx Hi!
1198764160 M * Flinx I did install X server and KDE on guest os, but i can run X server only with vesa drivers.
1198764228 M * Flinx With i810 drivers X server returns error: "Device not found"
1198764377 J * dna_ ~dna@p54BCD42F.dip.t-dialin.net
1198764398 M * Bertl Flinx: not unexpected, you need to give a bunch of capabilities and copy some devices to the guest
1198764417 M * Bertl Flinx: note that this makes the guest insecure, as it will have full hardware access
1198764450 M * Flinx Bertl: I copied ALL! devices to guest1 just for testing.
1198764474 M * Flinx Bertl: This is what i have in bcapabilities file:
1198764517 M * Bertl ah, and I forgot, you probably also need to make some /proc entries visible (/proc/bus and friends)
1198764647 Q * dna Ping timeout: 480 seconds
1198764662 M * Flinx Bertl: CHOWN
1198764663 M * Flinx DAC_OVERRIDE
1198764665 M * Flinx DAC_READ_SEARCH
1198764666 M * Flinx FOWNER
1198764668 M * Flinx FSETID
1198764669 M * Flinx KILL
1198764671 M * Flinx SETGID
1198764672 M * Flinx SETUID
1198764674 M * Flinx SETPCAP
1198764675 M * Flinx LINUX_IMMUTABLE
1198764677 M * Flinx NET_BIND_SERVICE
1198764678 M * Flinx NET_BROADCAST
1198764680 M * Flinx NET_ADMIN
1198764681 M * Flinx NET_RAW
1198764683 M * Flinx IPC_LOCK
1198764684 M * Flinx IPC_OWNER
1198764686 M * Flinx SYS_MODULE
1198764687 M * Flinx SYS_RAWIO
1198764689 M * Flinx SYS_CHROOT
1198764691 M * Flinx SYS_PTRACE
1198764692 M * Flinx SYS_PACCT
1198764694 M * Flinx SYS_ADMIN
1198764696 M * Flinx SYS_BOOT
1198764698 M * Flinx SYS_NICE
1198764700 M * Flinx SYS_RESOURCE
1198764702 M * Flinx SYS_TIME
1198764703 M * Flinx SYS_TTY_CONFIG
1198764705 M * Flinx MKNOD
1198764706 M * Flinx LEASE
1198764743 M * arachnist uhm
1198764763 M * arachnist please use phpfi.com or similiar page for pasting
1198764767 M * Bertl (please use paste.linux-vserver.org for everything longer than 3 lines) 
1198764804 M * Flinx oops, sorry
1198764805 M * Flinx got it
1198764840 M * AStorm give it /proc/bus
1198764905 M * Flinx this is what i have in vprocunhide-files http://paste.linux-vserver.org/10932
1198764939 M * Flinx all from proc/bus
1198765090 M * ard6 Flinx http://paste.linux-vserver.org/10933
1198765105 M * ard6 from a working intel GMA3000 system
1198765147 M * ard6 the device nodes you need is /dev/dri/*
1198765154 M * ard6 and /dev/input/event*
1198765168 M * Flinx ard6: Thank you! Will try now.
1198765169 M * ard6 and /dev/mem
1198765217 M * ard6 also take a look at http://oldwiki.linux-vserver.org/MoreUbuntu
1198765231 M * ard6 world-of-padman plays excellent on the GMA3000 :-)
1198765350 Q * AStorm Quit: ET calling home
1198765366 J * JonB ~NoSuchUse@0x5739c847.roennqu1.broadband.tele.dk
1198765466 J * hparker ~hparker@lonegroved20p183-d.cld.brightok.net
1198765517 J * doener_ ~doener@i577BA04A.versanet.de
1198765838 P * Flinx 
1198765932 Q * doener Ping timeout: 480 seconds
1198766344 J * AStorm ~astralsto@tor-irc.dnsbl.oftc.net
1198766457 J * Flinx ~chatzilla@86.57.216.113
1198766461 Q * AStorm 
1198766473 M * Flinx Guys, i did it!
1198766482 M * Flinx DRI is working perfect now!
1198766492 M * Flinx Thank you very much for help!
1198766539 M * Bertl you're welcome!
1198766576 M * Flinx ard6: Than you!
1198766613 M * Flinx You guys are geniuses!
1198766643 Q * JonB Quit: This computer has gone to sleep
1198767300 M * ard6 The big question however is : does world of padman work on your system? On my GMA865 it doesn't :-(
1198767334 Q * DLange Quit: Bye, bye.
1198767347 M * Flinx ard6: Toy want me try this?
1198767375 M * ard6 Flinx : your keyboard still has problems in the vserver 8-D 
1198767422 M * Flinx ard6: Sorry, i'm sitting in the darkness here
1198767428 M * ard6 Flinx : But about WOP: well, I am curious if it works... My workstation will be upgraded in 3 years because mine was 3 weeks to young in the last ugprade
1198767463 M * Flinx ard6: Ok, i will try
1198767472 A * ard6 secretly confiscated a workstation of somebody that doesn't work here anymore, and his was upgraded :-)
1198767562 M * Flinx ard6: this is WoP http://padworld.myexp.de/index.php?files ?
1198767583 M * Flinx ard6: Should i download from there?
1198767665 M * ard6 Flinx : Only if you were already going to try it out :-). It's only 550MB
1198767680 M * ard6 that's the correct url
1198767695 M * Flinx ard6: Ok, already downloading
1198767835 M * Flinx ard6: can i sent private messages to you? WoP is off-topic i think.
1198768363 M * Bertl yes, please do so ...
1198768831 M * arachnist how is WoP related to vserver? :>
1198768970 J * JonB ~NoSuchUse@0x535f65c3.kjnxx7.adsl-dhcp.tele.dk
1198769136 M * arachnist ard6: i used to play WoP on a p2 400 w/ 256MB of ram and gf4mx440
1198769201 M * JonB World of Pastry? a online baking site
1198769226 M * arachnist JonB: World of Padman
1198769234 M * JonB dont know it
1198769243 M * ard6 arachnist : WoP was just a side note, and a good test of DRI within a vserver 8-D
1198771088 Q * Flinx Quit: User is away.
1198771317 Q * JonB Quit: This computer has gone to sleep
1198771807 M * Bertl okay, off for now .. bbl
1198771811 N * Bertl Bertl_oO
1198772840 J * AStorm ~astralsto@tor-irc.dnsbl.oftc.net
1198773128 Q * AStorm 
1198773161 J * AStorm ~astralsto@tor-irc.dnsbl.oftc.net
1198773267 J * JonB ~NoSuchUse@0x535f65c3.kjnxx7.adsl-dhcp.tele.dk
1198773473 Q * Pazzo Quit: Ex-Chat
1198773534 Q * JonB 
1198773570 J * JonB ~NoSuchUse@0x535f65c3.kjnxx7.adsl-dhcp.tele.dk
1198774042 N * phedny Guest1394
1198774050 J * phedny ~mark@ip56538143.direct-adsl.nl
1198774140 J * ViRUS ~mp@p57A6FC86.dip.t-dialin.net
1198774383 Q * Guest1394 Ping timeout: 480 seconds
1198775018 Q * JonB Quit: This computer has gone to sleep
1198775269 J * bonbons ~bonbons@2001:960:7ab:0:2c0:9fff:fe2d:39d
1198775948 Q * yarihm Quit: This computer has gone to sleep
1198777584 J * JonB ~NoSuchUse@0x535f65c3.kjnxx7.adsl-dhcp.tele.dk
1198777597 J * rgl ~rgl@84.90.10.245
1198777600 A * rgl waves
1198777641 M * JonB rgl: bye bye
1198777806 N * DoberMann DoberMann[PullA]
1198777815 J * ntrs ~ntrs@vs079.rosehosting.com
1198778448 M * Wonka aaaaargh.
1198778472 M * Wonka /var/lib/vservers was remounted ro some hours ago.
1198778476 J * yarihm ~yarihm@84-75-119-160.dclient.hispeed.ch
1198778486 M * Wonka because iscsi had some glitch
1198778507 M * Wonka now the blockdev is back to rw, i think...
1198778524 M * Wonka but "mount -o remount,rw" thinks otherwise...
1198778536 M * Wonka going to have to reboot the whole machine :/
1198778554 J * Wonka_ ~produzier@25-199-103-86.dynamic.dsl.tng.de
1198778705 Q * Wonka Quit: need to reboot vserver host
1198778705 N * Wonka_ Wonka
1198778807 Q * pmenier Quit: Konversation terminated!
1198778926 M * pmjdebruijn Wonka: iSCSI on Linux is still crap
1198778932 M * pmjdebruijn we have disconnect issues
1198778992 M * Wonka we need something like iscsi here. one box has the storage, which cannot be put into the other, which has the processors.
1198779007 M * Wonka and it went well for several months now, iirc
1198779061 Q * larsivi Quit: Konversation terminated!
1198779241 Q * JonB Quit: This computer has gone to sleep
1198779750 J * JonB ~NoSuchUse@0x535f65c3.kjnxx7.adsl-dhcp.tele.dk
1198779819 M * rgl is AoE more reliable than iSCSI?
1198779858 M * JonB rgl: and FiberOverEthernet?
1198779899 M * rgl and why not use NFS instead?
1198779906 M * JonB "An alternative to iSCSI, the AoE specification is 8 pages compared with iSCSI's 257 pages.
1198779906 M * JonB "
1198779930 M * JonB rgl: NFS is files, not a block device
1198779953 M * rgl JonB, does it matter?
1198779963 M * JonB rgl: for some it does
1198779968 A * rgl does not known what is a block device *G*
1198780002 M * JonB rgl: NFS forces you to run NFS, getting a block device allows you to run any filesystem
1198780046 M * rgl JonB, but some FS are not distributed, can they still run over a block device?
1198780089 M * JonB rgl: a block device is a block device, it doesnt matter if it is a disk, a loopback file or a network block device
1198780094 M * pmjdebruijn comparing SAN to NAS is silly
1198780097 M * pmjdebruijn apples and oranges
1198780107 M * pmjdebruijn you can eat both though
1198780124 M * JonB pmjdebruijn: and make marmalade with both
1198780135 M * pmjdebruijn JonB: indeed
1198780153 M * pmjdebruijn NFS in Linux isn't _that_ great either
1198780169 M * pmjdebruijn if you can afford is, I can highly recommend FibreChannel
1198780183 M * JonB pmjdebruijn: or the brand new fiberchannel over ethernet?
1198780194 M * pmjdebruijn JonB: indeed "brand-new"
1198780196 M * rgl does a block device support multiple clients?  or this block device thing is just to move the disks to another box?
1198780196 M * pmjdebruijn aka immature
1198780209 M * pmjdebruijn iSCSI on Linux isn't crappy, it's just immature
1198780218 M * Wonka FCoE? is that a better concept than SCSIoE?
1198780221 M * pmjdebruijn you _don't_ want immature stuff for your storage
1198780235 M * JonB rgl: if you only READ it might support multiple clients
1198780254 M * pmjdebruijn iSCSI = SCSI over IP (and not Ethernet), that means it's has a bit more overhead, but it's routable
1198780255 M * JonB pmjdebruijn: which is why i stick to software raid and disks inside the machine
1198780261 M * Wonka JonB: what about cluster file systems?
1198780267 M * JonB Wonka: dunno, i just read about it the other day
1198780267 M * pmjdebruijn those are still crap too
1198780272 M * pmjdebruijn immature as well
1198780277 M * pmjdebruijn at least OCFS2/GFS2
1198780284 M * pmjdebruijn OCFS2 is very promising though
1198780288 M * rgl ZFS? :D
1198780295 M * JonB rgl: thats not a cluster
1198780297 M * pmjdebruijn ZFS is clustering? or is it?
1198780300 M * pmjdebruijn isn't?
1198780306 M * JonB not to my knowledge
1198780320 M * rgl they said they were incorporating the lustre stuff into ZFS to make it clusterable
1198780325 M * rgl (they sun)
1198780332 M * pmjdebruijn that could be
1198780349 M * pmjdebruijn though lutre clustering is a bit different from OCFS/GFS style
1198780352 M * JonB rgl: i do not think they released anything yet
1198780366 M * pmjdebruijn it's a very different approach, which may not be suitable for all purposes
1198780376 M * pmjdebruijn especially where POSIX semantics are of importance
1198780387 M * rgl JonB, ah yes.  it some vaporware they were talking about.  though, they've both the lustre software and all *G*
1198780393 M * pmjdebruijn that's the nice bit about GFS/OCFS, they practically fully POSIX compliant
1198780395 M * rgl err s,both,bought,
1198780448 M * rgl pmjdebruijn, so you are using a block device to serve a single server, correct?  what FS are you using?
1198780499 M * pmjdebruijn we use FibreChannel (with QLogic HBA's/switches and HP storage gear) with XFS
1198780525 M * JonB rgl: i see it more like an annouincing that they are going t make that
1198780539 M * JonB rgl: i think thats pretty common in the commercial world
1198780552 M * rgl JonB, yeah, vaporware :D
1198780556 M * JonB pmjdebruijn: how many data are you storing?
1198780556 M * pmjdebruijn rgl: that means each server has it's own partition on the storage as, a single partition/block device isn't accessible by multiple servers
1198780562 M * pmjdebruijn JonB: lots... :)
1198780569 M * pmjdebruijn JonB: tens of TB
1198780583 M * JonB pmjdebruijn: okay, i only have about 1.6TB
1198780589 J * Wonka_ produziert@chaos.in-kiel.de
1198780597 M * rgl pmjdebruijn, ah.  so, iSCSI and AoE is really just replacing the cable to the disk.  the rest is +- the same:D
1198780617 M * rgl (replacing the cable to disk with an IP/ethernet connection/cable)
1198780626 M * pmjdebruijn sooner or later iSCSI will mature, and by that time it will be a good replacement for most uses of FibreChannel
1198780628 M * Bertl_oO AoE works fine here, btw, and IIRC Aiken tested that extensively
1198780639 M * pmjdebruijn though there are mature iSCSI HBA's available on the market
1198780647 M * pmjdebruijn it's just Linux' software implementatino which is immature
1198780656 M * rgl whats HBA?  *G*
1198780667 M * pmjdebruijn rgl: host bus adapter (i.e. hardware controller)
1198780670 J * ntrs_ ~ntrs@vs079.rosehosting.com
1198780693 M * rgl ah!
1198780714 M * JonB i think that AoE, iSCSI only introduces more complicated stuff between the filesystem and the raw hardware
1198780803 M * JonB has anyone ever tried experimenting with having a filesystem interface to the data? rather than a block interface? I can see the filesystem being an advantage to virtual machines like vmware
1198780831 M * Bertl_oO hmm?
1198780896 M * nebuchadnezzar JonB: do you mean, using a file image instead of a raw partition or logical volume ?
1198780921 M * rgl naive question: why does one want to use NAS instead of attaching the HD into the machine through a local interface like PCI?
1198780926 M * pmjdebruijn JonB: filesystemen interface to the data????
1198780929 M * pmjdebruijn JonB: that's NFS/SMB
1198780941 M * JonB nebuchadnezzar: no, more like requesting a file from the underlying "hardware" or emulation of hardware
1198780946 M * pmjdebruijn rgl: easier management
1198780956 M * pmjdebruijn rgl: centralisation
1198780971 M * pmjdebruijn rgl: plus, RAID 5 on 10 servers means at least 30 disks
1198780996 M * pmjdebruijn in a big enclosure you have RAID 6, over 26 disks, effectively using 24
1198781006 M * pmjdebruijn which means you have _much_ less overhead
1198781015 M * JonB pmjdebruijn: yes, you are probably right that would work, but NFS is networked, i was thinking local only
1198781041 M * pmjdebruijn my advise... seriously consider investing in HBA's for either FibreChannel or iSCSI
1198781056 M * pmjdebruijn if you decide to use NFS, make sure you test it thoroughly
1198781072 Q * ntrs Ping timeout: 480 seconds
1198781075 M * pmjdebruijn we have had issues with load (with kernels as recent as 2.4.22)
1198781082 M * rgl pmjdebruijn, how do you have less overhead?   does the iSCSI where the disks are have a dedicated controller and no normal MB and OS?
1198781089 M * pmjdebruijn no
1198781101 M * JonB pmjdebruijn: you can get that in a machine as well, like the 	
1198781102 M * JonB Sun Fire X4500 Server
1198781111 M * pmjdebruijn rgl: 10 x 3 disks RAID 5 = effectively 10 disks of usable space
1198781131 N * Wonka W0nka
1198781131 M * pmjdebruijn rgl 30 disks of RAID 6 via iscsi/fc = effectively 28 disks of usable space
1198781131 N * Wonka_ Wonka
1198781140 M * pmjdebruijn anyway
1198781142 M * pmjdebruijn I'm off
1198781143 M * pmjdebruijn bye
1198781205 M * rgl I'm not understanding your reasoning.  RAID6 should use more disks than raid5;  how does that matter to the NAS case?
1198781222 M * JonB rgl: raid6 is safer than raid5
1198781226 M * rgl I'm just discussing NAS vs local disks.
1198781231 M * JonB rgl: okay
1198781264 Q * W0nka Quit: back
1198781296 M * rgl I could thing it has less overhead if the NAS box has some dedicated stuff to connect to the disks (eg, no bottleneck in the local bus).
1198781371 M * JonB rgl: i think the advantage is that you can add another NAS box, but adding more disks to a server is not that easy
1198781454 M * rgl ah, indeed. :D
1198781513 M * JonB rgl: and you can hire a guy to run the NAS and let others run the servers
1198781545 M * rgl true true :D
1198781565 M * JonB then the server guy can just request X amount of GB storage
1198781583 M * JonB and let the NAS guy worry about keeping it running, and possibly backup as well
1198781607 M * rgl at at first glance I'm not buying iSCSI, it seem so inactractive regarding AoE *G*
1198781707 M * JonB rgl: AoE are over ethernet, iSCSI are over IPv4 or 6, and thus needs to be more complicated
1198781722 M * rgl humm how can the NAS guy known how to backup?  isn't NAS a dummy thing?  eg, you can't make a stable backup when someone is still using the device.
1198781747 M * JonB rgl: with some NAS devices you can say "take a snapshot now"
1198781749 M * rgl JonB, indeed.  I don't see why one would need to put a NAS over IP ;-)
1198781756 M * JonB and then you can backup the snapshot
1198781777 M * rgl how can that work?
1198781787 M * JonB how can what work?
1198781798 M * rgl eg: the databases would not be ok, or will they?
1198781828 M * rgl I mean, suppose you have a database like MySQL, and you take a snapshot, can you rely on the snapshot?
1198781856 M * JonB rgl: well, suppose like this, you tell the server to shutdown the services, signal to the NAS to take the snapshot, and then restart the services
1198781863 M * JonB downtime is seconds
1198781888 M * JonB then you can just backup the snapshot even if that takes 23 hours and 59 minutes
1198781895 M * rgl yup, that way it seems to work :)
1198782181 M * JonB rgl: i am thinking of doing the same with LVM, and vserver guests
1198782269 Q * yarihm Quit: This computer has gone to sleep
1198782306 J * yarihm ~yarihm@84-75-119-160.dclient.hispeed.ch
1198782410 Q * JonB Quit: This computer has gone to sleep
1198782488 Q * ema Quit: leaving
1198783781 Q * hparker Ping timeout: 480 seconds
1198784215 J * JonB ~NoSuchUse@0x535f65c3.kjnxx7.adsl-dhcp.tele.dk
1198784293 J * hparker ~hparker@lonegroved20p183-d.cld.brightok.net
1198784295 J * harry ~harry@d54C363EF.access.telenet.be
1198784525 Q * dna_ Read error: Connection reset by peer
1198784550 J * dna_ ~dna@p54BCD42F.dip.t-dialin.net
1198785537 Q * hparker Ping timeout: 480 seconds
1198786002 Q * meandtheshell Quit: Leaving.
1198786685 J * hparker ~hparker@lonegroved20p183-d.cld.brightok.net
1198787843 N * AStorm Guest1407
1198787847 J * AStorm ~astralsto@tor-irc.dnsbl.oftc.net
1198787863 Q * Guest1407 Ping timeout: 480 seconds
1198788193 N * AStorm Guest1408
1198788197 J * AStorm ~astralsto@tor-irc.dnsbl.oftc.net
1198788522 Q * Guest1408 Ping timeout: 480 seconds
1198789003 Q * yarihm Quit: This computer has gone to sleep
1198789066 Q * hparker Ping timeout: 480 seconds
1198789267 J * hparker ~hparker@lonegroved20p183-d.cld.brightok.net
1198789961 J * yarihm ~yarihm@84-75-119-160.dclient.hispeed.ch
1198789996 Q * hparker Ping timeout: 480 seconds
1198790452 M * Bertl_oO okay, off to bed now ... have a good one everyone! cya!
1198790458 N * Bertl_oO Bertl_zZ
1198790873 Q * dna_ Read error: Connection reset by peer
1198790899 J * dna_ ~dna@p54BCD42F.dip.t-dialin.net
1198791425 Q * dna_ Read error: Connection reset by peer
1198791451 J * dna_ ~dna@p54BCD42F.dip.t-dialin.net
1198791484 J * ntrs ~ntrs@vs079.rosehosting.com
1198791872 Q * ntrs_ Ping timeout: 480 seconds
1198792213 Q * JonB Quit: This computer has gone to sleep
1198792634 J * Aiken ~james@ppp121-45-246-228.lns2.bne4.internode.on.net
1198793000 Q * bonbons Quit: Leaving
1198793108 Q * ard Ping timeout: 480 seconds
1198793277 Q * harry Ping timeout: 480 seconds
1198793458 J * dna__ ~dna@p54BCD42F.dip.t-dialin.net
1198793468 Q * dna_ Read error: Connection reset by peer
1198793588 J * harry ~harry@d54C363EF.access.telenet.be
1198794865 Q * dna__ Quit: Verlassend
1198795394 Q * _er Remote host closed the connection
1198796204 Q * rgl Remote host closed the connection
1198796687 Q * yarihm Quit: Leaving
1198799260 J * ard ~ard@gw-tweakb16.kwaak.net