anybody have multiple vps'es and notice network disconnects lately?
aka I have 4 vps'es .. 2 are able to talk to the v4 gateway, 2 are able to talk to the v6 gateway .. I'm going to guess something changed between 22:30 yesterday and 01:32 today based on my logs
this probably isn't gonna paste right...
01:06:222014-01-15MenuhopDOWNCRITICAL - 174.136.107.10: rta nan, lost 100%
meh
0506-0533 utc (today), looks like our monitoring system couldn't hit that vps
oh wait, ignore that. that was on the 15th.
for what it's worth,  one of mine seemed to be unavailable for about 45 minutes last night.
wait,  wrong provider.  my ARP vps's seem fine
I guess I'll have to blame it on my non-prescription drugs?  lol
lol
ahh, it wa a digitalocean vps that was unavailable.
i've had some weird intermittent disconnects
never enough to trigger monitoring
err, trigger notification
I think this was a maintenance that went wrong for my vlan only, if nobody else has issues
{0,2,3,5}.v.freedaemon.com .. all are reachable via one address family, prior to last night, all were reachable via both address families
huh.
;-)
and of course the ones that are only on v6 can't reach the ones that are only on v4
so my vlan is effectively two network segments
there's one of the blips i mentioned, haha
12:19:46 ::: m0unds [~m0unds@2607:f2f8:af18:2::e] has quit [Ping timeout: 276 seconds]
yep
I love sprunge!
http://sprunge.us/jDea
i have to admit
sprunge sounds like something you'd see in a porno
:V
lol, also not easy to remember (or so I find)
which is why the company's private clone of sprunge is simply {company}pastebin.domain.com
oh my bad, pastebin.domain.com
my buddy made an erlang webapp that passes stuff via zsh to the pastebin attached to the webapp
also handles url shortening and image fetch/shortening - so if you paste an image url into irc, it'll retrieve the image, store it temporarily on s3 and serve it via shortened url
"webapp that passes stuff via zsh" that sounds pretty damn risky :/
^
That totally sounds like an Erl Web app.
uses an API key and stuff
heh
but i do know you can pipe output from commands (top -m io etc) and have it display as text
https://twitter.com/arpnetworks/status/427969490203066369
TWITTER: We will be performing an emergency reboot of our IPv6 router at 19:00 PST tonight (Tue Jan 28 01:00:52 +0000 2014, retweeted 1 times)
"emergency" and scheduled rarely go together :p
@twitter -r 427969490203066369
brycec: Successfully retweeted arpnetworks (427970979109691392): We will be performing an emergency reboot of our IPv6 router at 19:00 PST tonight (Tue Jan 28 01:00:52 +0000 2014, retweeted 3 times)
wow, earlier reboot than expected! :P
well, it's like in 2 hours...
pretty close IMO :)
up_the_irons: just a reboot? or upgrade too?
just reboot
it needs a 2nd interface from our s7.lax router
When does ARP get a redundant IPv6 router?
ah
yeah no immediately plans for 2nd IPv6 router
Get a cheap, used DL380 or similar, a little CARP action and BAM
if it were that easy, it would have happened already
:)
actually, s7.lax can do full IPv6
in hardware
but the configuration dance and playing nice with existing OpenBSD IPv6 router, that is the hard part
brycec: s/380/360/
but carp complicates matters
often non redundant solutions can be more reliable with decent hardware
heh
My experience with redundant routers extends to pfSense
Which was really pretty simple
you can get into lots of partial failure situations
like when there were router issues the other night, that was partial failure rather than full failure
heh I did have a bit of split-master fun today. But prior to that... :p
you have to think about things like comlpexity of debugging
like if you have an  ethernet card giving issues, in a redundant situation, which one is it
but yeah, if was to have a cheap secondary router, a dl360 is 1u, dl380 2u
and shouldn't need more than 1u
they both can do dual cpu etc, not that openbsd makes use of that
I just pulled an HP DLxxx number out of my butt as an example of something that's cheap and readily available.
ahh right
notification gave me advanced warning to move my tunnel stuff over to my bigger vps
haha
s/bigger/non-ARP/
<m0unds> notification gave me advanced warning to move my tunnel stuff over to my non-ARP vps
I assume
no
s/as/pre/
<brycec> I presume
bigger being bigger at arp
i think arp is using supermicro normally?
which have much more confusing model numbers :)
i planned on doing it around when it was gonna get rebooted
up_the_irons: does that mean ipv6 is going to be gigabit?
mercutio: if you're already on gigabit ports,  ummm.. yeah i think it might actually (haven't really even thought about that)
 hahaha
i was just curious more than antyhing
i still see ipv6 as being beta'ish
"second class"
yeah it's a bad attitude to have in some ways, because the more people who think that way the more it becomes that way
heh
The native IPv6 was a big reason I moved to ARP
A little sad that it doesn't get the same love as v4
it was secondary reason for me
i was tunneling with he.net prior
until more eyeball networks support ipv6, i don't really care one way or another
but i got over ipv6 :)
it doesnt' really help having ipv6 support when the situations it'd be useful don't have ipv6
yea
and in a lot of cases, it can break stuff for client devices
like with ipv6 and static ip, you'd be able to remote desktop into someone's computer if they oppened it up
but it's easier to talk smoeone through doing a port forward then setting up ipv6  even if their provider supports it
yeah
can't imagine having to tell someone an ipv6 addr over the phone or something
haha
yeah that's one of the current issues with ipv6
should every user host their own dns?
then the differing implementations at the cpe sucks too
as soon as you have all these ip addresses, and want to do dns, things get tricky
on legacy ipv4 connections there's susually just a single or a few ip's to give a name to
and try running a mail server without reverse dns...
yeah
and i don't think most isp's have user-modifiable dns
actually  arp did hah
i dunno about ipv6, but with ipv4 you can easily change your reverse lookup
but try doing that on a normal net connection...
yea, you can w/arp, but i've found i don't always get an rdns when i query an ipv6 addr
fortunately, with most isp's you can email a request for reverse dns
as long as you have a static ip alloc, yeah
does arp let you set reverse dns?
i have rdns on mine
yeah
err for ipv6
yes
yes
2607:f2f8:af18:2::1
1.0.0.0.0.0.0.0.0.0.0.0.0.0.0.0.2.0.0.0.8.1.f.a.8.f.2.f.7.0.6.2.ip6.arpa domain name pointer lax-gwy.m0unds.net.
heh cool
mercutio: i'm trying not to treat IPv6 as "beta-ish"; first class IPv6 has been a goal of mine since the beginning; but traffic levels used to be so low it was hard to prioritize
brb
up_the_irons: well that's good to hear, now get evyrone else onboard :)
oh so ipv6 is down now?
not that i was using it normally, was just curious if facebook has reverse dns
oh right it's just on 19:00 PST isn't it :)
yup
mercutio: yup 19:00 :)
it's back up now
sucking down the bgp routes
routing not getting anywhere
but gateway responds
ahh ok
openbgpd is slow at loading route tables i've noitced
at least ipv6 is small
heh just rerouted IRC services over the v4 VPN link temp
hmm routing still getting nowhere
oh facebook works
ahh kame.net works now too
wow facebook has reverse dns
BGP is slow eh
on all of their hops!  even their routers
so do ntt
www.kame.net don't for the web server itself
google only have web server with reverse dns, none of their routers
|- scruffy.entropynet.net[46X] -------------- | Users:     7
server didn't even netsplit :)
didn't seem like i lost connection either
<.<
that seemed short/quick
yep
that's what she said
lol
I've had more outages with my tunnelbroker.net tunnel than with arpnetworks
he.net?
One in the same