/buffer 5
erm
fail.
its ok cmeiklejohn
we forgive you
thank god
speak for yourself.  I don't forgie him for such an egregious error.
in fact..  I was almost debating if I should tar and feather,   or mock and ridicule..
/win -12
 /lose
/fail
well.   one thing is for sure. ARP runs a decent network.  if we have time to sit back and be silly on IRC, then the provider is doing something right. :P
haha
indeed jpalmer
are arp VM's running under KVM?
cshobe: yes, it should be all outline in the FAQ
"This changes the kernel so it can modify itself when it is run under a hypervisor, potentially improving performance significantly over full virtualization."
I reckon I'll be enabling that then
I thought so
jpalmer: LOL, I like your assessment
cshobe: but we don't use a hypervisor so don't enable that ;)
cshobe: full virtualization with hardware support (Intel VT) is superior to paravirt w/ hypervisor
+1
my laptop is not quite new enough to have VT, and I really miss it
up_the_irons: eh, /proc/cpuinfo shows 'hypervisor' in the cpuflags?
cshobe: it lies
up_the_irons: fair enough, got any input on NUMA support?
no comment on NUMA
also, linux seems to want to use /dev/vd* now rather than /dev/sd*
is it trying to be too clever for it's own good?
cshobe: use vd*, b/c the arch is virtio, not scsi
virtio = faster
You can pry my emulated ATA from my cold dead hands!
trying to remember... physical hosts are linux, yes?
yes
I thought I heard that some were freebsd?
nope
quick question are the VPS accounts limited to 10mbit incoming ? (doing a quick fetch from another server of mine which does 100mbit out) getting at most around 900kBps during this transfer test
but course it goes up and down between 500 and 900
It looks like traffic into the box is limited to around 400-700 on average, but traffic of the VPS does a pretty steady 1-1.25MB/sec
but couldn't find anything on the website that listed official uplink speed
well did find "Each VPS shares a 100 Mbps uplink with other VPS's running on the same host machine." , but doesn't say if there's limiting (which there seems to be to 10mbit)
Also that's uplink not the other way. Hm. Where are you downloading from?
boba: was testing both ways , fetch on the arp network from a dedi I have with a 100mbit uplink. Then I used wget on the dedi to pull it back accross
 from dedi to arp vps , round 400 to 900 up and down during the 100mb test
 from arp to dedi a steady 1.15-1.2MBps
was going to do the test again on another VPS I have, but the dedi is the one that I have that's over 10mbit
so would certainly utilize the max available from arp either direction
What's the routing like between arp and your dedi?
like tracerouting ?
also I tried the test file from arp's page to the dedi which was half the speed, but sec I'll get a traceroute for you
http://pbin.be/show/mSvhUEN7Fz8QqWUOEuEg/
performed both ways ( just in case they're routing differences)
though seems like both locations are connected by he's backbone
I can pull upwards of 3-4 MB/s from another VPS I have in the Northeast U.S., so...
http://208.79.90.98/100mb well feel free to grab it from the arp vps
2011-12-05 20:10:40 (11.1 MB/s) - `100mb' saved [104857600/104857600]
hrm...
sec
bout 2.5M/s max now (flushed the tc rules which shouldn't have been affecting that IP but they're all cleared out)
bout 3.1Mbps max into the vps
That sounds a little more reasonable ;)
k, guessing it's just my TC rules then on the dedi that were limiting strictly to 10mbit or less
had to revert to using tc, cuz the latest update for SolusVm removed the rate limiting feature and they didn't bother to tell any customers
(and month later they still don't have an updating putting that feature back in)
well that's one less thing to test (I'm comparing a cheap bitcable account, to the arp vps, and might test out rootbsd)