hm.. did anyone else mention not being able to log in to VNC on kvr13 recently?
I get "connected" but no screen (and depending on the vnc viewer, no password prompt..)
I suppose I should hard shutdown
that got it ..
vnc -> tcp connect direct to the userland kvm process, so .. if kvm has issues serving tcp, then that will suffer
'hard shutdown' -> kill kvm process, then 'boot' -> 'start kvm process'
toddf: yeah
ARP what did you do to me?! I've been using OpenBSD for a few weeks, and here I am now stting in front of a debian box and thinking.... "Damn... Linux sucks..."
:(
do you mean you lost your openbsd install? :(
nope
you converted to using debian then? :P
but I had to use this debian box for someone
ah
my vpses are still kicking obsd :-p
:)
what I mean is after years of only linux, I realise now that it's a little messy
maybe 'sucks' is an over reaction though
yeah, I definitely like how organized things are (in general) on the BSDs
tinono: it should follow that things are organized and uniform in projects where all things are controlled in one place by one set of developers... ;-)
greets - is there a way to change the VNC password? I couldnt see how
and yes, OpenBSD makes much more sense
not presently, support@ ticket to bump the priority of changeable passwords across the board .. I suspect it will eventually happen, just subject to up_the_irons time etc ;-)
ok
i'm actually glad to finally get a decently priced bsd server
i'm migrating off linux now
I looked around for a while, there are some places that are slightly cheaper by a few pennies but nothing beats the remote access and instareboot capabilities
yes, i dig the remote vnc console
serial console also..
i havent tried that yet
but the host is running linux kvm right? do you have any idea how it handels when memory is tight?
my last vps was a virtuoza, and it would kill processes
send a ssh pubkey to support@ and get ssh access to shutdown/boot/secure vnc tunnel/connect to serial console/etc
read the signup page, arpnetworks does not overcommit memory
thats one of the reasons i signed up, great in theory
bharatak: sounds like your old provider overcrowded the servers
bharatak: yeah, RAM is not over-committed.  Each host machine has 32GB of RAM, and I put up to 24GB of VMs on it, leaving the remaining 8GB for the OS + possible RAM stick failure (thereby leaving only 28GB left, so you need the buffer)
I've had a stick fail once
RAM is one of the few components I'll spend a premium on.  I buy a ton if it,  and have never had an issue with a stick from crucial.com,  but there are several good brands
jpalmer: i buy from crucial too, good experiences.  however, i have an 8GB kit from my last purchase that is actually bad.  need to RMA it.  But at least it was the bad right away, and not 3 months down the road after a production box is deployed ;)
oh man, thats the worst. yea.
random corruption and./or crashes.
up_the_irons: I've *never* had to RMA from crucial.  I'd be interested to hear how it goes.  if it's a pain, or if they are quick about it and willing to cross-ship,  etc
jpalmer: yeah, i'll let u know
danke
jpalmer: every manufacturer has ram go bad eventually.
jpalmer: I mean shit, I have to go hunt down bad ram this weekend :/
really hoping just turning off ECC will be enough of a test, I don't want to sit there waiting for memtest to churn through 4gb 4 times (worst case, I guess)