^ @cole

@cole You'd be great for this position

we also require that you be able to not lose your GPG private key

@FalconMomot image not found.

@Jacob my Linux skills are duller than a spoon now.

@FalconMomot It's not lost!

@cole it's a one-eyed three-horned monster

It's on another machine!

@pauska Nope, DPM has it's own server.

The other day I couldn't remember how the hell to get the RHEL version.

@cole Your ops experience is more important.

You can learn the nix stuff quickly

if it isn't uname I don't know either but I bet I can find out before anyone notices.

@ChrisS is it also uber slow to remote in to?

My brain is just so fried that I forget simple stuff.

ours use a minute or two to clear out enough memory for RDP to work...

@cole It's okay, I always forget which box I'm on so I yum on my Debian box and apt-get on my CentOS one...

@Jacob sadly, Seattle is not in my near future.

@cole cat /etc/redhat-release

@pauska No, the RDP connection is as quick as any... The application is slow, but not the connection

@Jacob I know now but at the time I was like "uh....potato."

@cole You don't need to memorize it

@cole cat: spud not found

You just need to know how to find it

@NathanC thank you, I needed that chuckle.

I think we need a canonical backup solution post.

I bit my tongue /tableflip

@cole Oh right. HE showed up and then bailed.

@ewwhite Just use xcopy. It'll be fine.

@Wesley yas

@Wesley seriously... I don't know what to do for backups these days

didn't we have a question recently where someone wanted to yank out drives from their rad1 for "backups" ?

@Wesley robocopy

@pauska we did "broken mirror" backups for years at the $TELCO I worked for.

@RyJones how many times did you end up with URE during rebuild?

Every night, yank out half the mirrors and put in new drives, rotate the mirrors through storage

@pauska Yeah I believe so...and I think it was closed/downvoted to oblivion because it was breaking horribly on them

Coworker who came in 45 minutes late - left 30 minutes early.

@pauska Never

@RyJones ..

36GB 15k disks?

@ewwhite I've been doing a zipped backup to a NAS which is backed up to an external USB drive to be taken off-site

@pauska ah these were older and smaller. en.wikipedia.org/wiki/Tandem_Computers

Although that's only a suitable solution for smaller setups...

they were probably 250 meg drives

@NathanC do you trust your NAS?

@ewwhite Well, I also have two in a HA cluster, so yes.

One a year, I do recovery on consumer NAS devices... usually Linux+XFS

Really? We're still getting Heartbleed questions?

The IT director of said $TELCO would test disaster recovery by walking into the data center and pulling power cords / flipping breakers for systems.

We'll be hearing about Heartbleed and Shellshock for the next year or two.

night punks

@RyJones I've heard of that...that practice even has a name that I can't quite remember

@RyJones I do that when I setup new hardware... Tell ya real quick if you've got it right or not.

@NathanC I don't recall. He'd make us pull out the manuals and walk through the recovery steps to prove they worked.

I have a Disaster Recovery test tomorrow.

he threatened to bring in his kids to do the recovery, but thankfully he never did

@cole eh, my doctor told me I'm fat and I'm going to die, so I'm not going back to him any more

Anyone familiar with ISC DHCP?

option MotoAP650 code 189 = string;
option MotoAP650 "172.30.27.21";

does that look right?

@ewwhite Um.... no? But I'm not that familiar with ISC DHCP. I am familiar with parsers, though, and that would give a parser fits if it was valid. So it probably isn't.

that's what I'm thinking... it looks off

But I guess it may be valid.

jurjenbokma.com/ApprenticesNotes/LWP_DHCP.html

wouldn't you add option MotoAP650 code 189 = string; in the root config

and then the option MotoAP650 "172.30.27.21"; in the subnet?

or something like that

Maybe?

that's how you do it on windows servers.. first define the option (string, bit etc) and then configure the option on the scope(s) or globally

I'm trying to convert this to a Windows 2012 DHCP failover cluster

and get it off of Linux

I feel your pain

ISC DHCP is so ridiculously complicated for such an easy protocol

My DHCP config for this company is 415 lines

and it's serving DHCP for 12 locations... (which is probably a bad idea)

# Plant 3
subnet 172.30.23.0 netmask 255.255.255.0{
        authoritative;
        pool {
                range dynamic-bootp 172.30.23.100 172.30.23.200;
                #failover peer "dhcp-#failover";
                deny dynamic bootp clients;
        }
        option routers 172.30.23.1;
        default-lease-time 72000;
}

you were right

I ALWAYS AM

@ewwhite Yeah, that seems right.

@ewwhite All depends on 1) The application, and 2) the filesystem

@pauska Right, the declaration (first line) must be in the root of the config. The setting can be either in the root (for a global setting) or within a subnet.

@Wesley but is there a de facto small business backup solution?

@ChrisS which means I need to figure out if it was meant to be a global setting across all 12 locations.

@ewwhite For very small locations, Windows Server Backup.

and then duplicate that functionality in Windows

Right, nothing to it.

(so glad DHCP is clustered in 2012)

Meh - wish MS and ISC could have come to some agreement on the failover DHCP functionality so they were interoperable. Or (possibly) better yet, open source their changes and merge back into a single solution...

I wonder how much of the ISC code is still in the MS DHCP service

Oh, there's a history there?

@ewwhite isn't there always?

@ewwhite Okay, I made a mistake. It depends on the OS primarily, then the apps involved, then the FS involved. The the RPO/RTO too

But as for OS, de facto is hard. In Windows the built in backups are okay, but not application consistent out of the box for things like Exchange or SQL Server, and possible other apps that have in-memory / journaling styles of data flow.

I usually end up making some mothership script that dumps, exports, or otherwise handles all the dorky apps that a small business uses before then kicking off a windows image backup or a robocopy script

@ewwhite nearly everything in the MS tcp/ip stack was derived from BSD+ISC

@ewwhite The original Windows DHCP and DNS server were modified ISC servers.

ah

afaik the DNS server in Windows 2000 Server was 100% compatible with BIND

@pauska Originally, correct

@ChrisS hence "was" :)

You could probably find traces hidden somewhere, but it was all BSD code so..

@pauska I know you know, just want to be clear that much has been completely rewritten over time.

or well, BSD licensed

If you value your privacy, I advise against taking a taxi in New York City.

Yeah, the original IP stack was mostly lifted from BSD 4.x

SO BSD is everywhere

except for my Macintosh :)

yeah, no... OSX is still about 50/50

OS X is built upon the Mach kernel, which derives from BSD through nextstep etc

Mach is not from BSD-land.

Mach was developed by CMU, Nextstep grafted Mach into a BSD userland. OSX put a pretty GUI on it.

Mach was initially hosted as additional code written directly into the existing 4.2BSD kernel, allowing the team to work on the system long before it was complete. Work started with the already functional Accent IPC/port system, and moved on to the other key portions of the OS, tasks and threads and virtual memory. As portions were completed various parts of the BSD system were re-written to call into Mach, and a change to 4.3BSD was also made during this process.

but yes, it started out on VAX or something like that

but I gotta head home, later

"About 16% of unique IP addresses that connect to CloudFlare do so via IPv6 (note: that calculation takes only the first 8 bytes as unique in any IPv6 address connecting to our network)."

Yeah... That graph makes it look like the kernel is some kind of "continuation" of the BSD kernel... Mach is a microkernel design where BSD is an extensible monolithic design. Mach was developed with a BSD userland to allow for rapid development of a usable/testable system, but relatively little of the Mach kernel code is shared with the BSD kernel.

@MichaelHampton Only the first 8? That can't be right..

@ShaneMadden That covers /64

@MichaelHampton Oh, 8 bytes, not bits. It must be Monday.

Now if only CloudFlare would get their OCSP responder working.

Seems reasonable since most IPv4 addresses are going to be coming from behind NAT, so it would be comparable to consider a IPv6/64

Reasonable enough, but it will undercount

I wonder how many devices the average person uses these days... Also how many users the average computer has these days..

4-5 IP devices?

I use 5 (work laptop, home laptop, home workstation, phone, tablet). But I share the table with the wife. She's at 4 (work computer, home laptop, phone, tablet)... So 4-5 sound about right...

@ChrisS im at 9 if you include local servers

@MichaelHampton Still working to get my TV remote unlocked by The Cloud

@ewwhite Good luck with that!

@MichaelHampton It's been an issue since June, I think.

maybe July

I just got a response from the product manager

Thanks for contacting me and apologies for the confusion and frustration.  Your case is definitely unique in that you removed your original remote from your account and replaced it with a new one, that we subsequently locked.  I can see how that would cause confusion for my team.

Regardless, I can see in our system notes that your remote is approved for unlocking and escalated our Level 2 support who can perform that.  I'll give them a nudge to ensure it happens as soon as possible.

I'm like: Thank you for the response, Ian. Will I need to take any additional action on my part, or will the unlock flag be handled entirely by your team?

Should be all on my end.  Once your remote is unlocked, you should be able to use it as though none of this ever happened.

Signed - Director, Customer Experience, Logitech Harmony Remotes

how is your tv remote working through the cloud?

Logitech Harmony remotes all access a hosted application that maintains remote state, settings, etc.

and one of my remotes failed...

so I did an RMA

but bought a new one in the interim

Logitech told me to throw the old defective remote in the trash...

whoa, RMA, I bet they've never had an RMA before. I can see why they got confused

when they "decommission" a remote, they disable it from their central system...

@ewwhite I had a nice backlit Logitec keyboard that had a defective space bar. They said "Meh, just keep it"

rendering it useless and unable to download software/firmware...

Sent me a new one

so I toss the bad one..

register my new one

Logictech disabled it

so the wrong device was bricked

@ewwhite Jackwagons.

so I had to produce the defective one and send it back to them

and the excuse I've heard for the past 8 weeks was, "The system won't allow us to unlock your remote"

@ewwhite They probably never even wrote code to unlock remotes.

as though "The System" acts on its own will

Ugh. I have such a headache.

@Wesley i hear that

Oh, hey, Apple finally released a bash update

@Wesley Let me rub some semen on it for you

What the fuckity fuck

One of our systems at a client site has been offline for like 3 weeks now, and I finally get an email after chasing them and chasing them and chasing them

> So sorry that I couldn’t get back to you sooner. I have just had a look on the server for you. It seems like the issue is stem from the fact someone has changed the network drive letters around on FINANCE118 which is a virtual server. Unfortunately, I can’t fix it myself.

> And I believe this is caused by the fact that the boys from IT helpdesk were moving the VM’s around last week. I will email them and let them know. Hopefully we can get this fix ASAP. I will let you know right away when it’s fixed.

Some dickwad changed the drive letters on an SQL Server and didn't think to test anything or even notify anyone that they had changed the drive letters

@MarkHenderson the shit

So @MarkHenderson - do you do IT consulting?

@cole We're sort of like SAP. We have a large system, most people chose to host with us but sensitive clients choose to self-host, in which case they pay us to support the software, but we require them to maintain their own hardware

@MarkHenderson ah ok

In this case, it's a government installation

@MarkHenderson just like produce software!

@ewwhite And like your produce installations, when this shit goes down, people sit around twiddling their thumbs

heh

They fell back to manual paperwork for this period, and whne the system is back online someone is going to have to go and back-enter it all

Not my problem, or my fault, but you bet I won't hear the end of it

Just imagine how terrible things must be at government agencies responsible for produce!

In fact I have an in-person meeting in 90 mins with them

@freiheit they go pretty well... The government breaks produce up into regions and contracts the work out to normal produce firms.

those firms compete for the contracts, but luckily my clients all get the government/military business..

just sayin'

@ewwhite I was thinking more like FDA Produce Inspectors‌​...

Oh, I get some of that, too...

USDA inspections.

heavily regulated...

back to my domain controller installation

Found an upside to shellshock: we're finally working fast on retiring this old Dell 2950 running RHEL4.

I now have a lady who is insisting that our quote template is no good, because she is taking the PDF, printing it out, annotating it by hand, re-scanning it in and then emailing it to the customer and "It looks pretty poor"

@MarkHenderson You make me so happy that I don't have to work directly with users.

@freiheit I'm not meant to be but 2 years ago our person who does deal with them went to work in the mining sector and his replacement only started yesterday

I think her exact wording was "It looks like a 4th grader did it"

Which is pretty insulting

@MarkHenderson Maybe you should work on being scarier

@MarkHenderson Because your 3rd grader would do it prettier than that? (they'd fill it out online and then print it)

@freiheit Because it looked fucking fantastic and we usually get compliments on the mix of useful functionality combined with still being visually appealing

Until she printed it out and ran her red biro all over it and scanned it back in with a $2 scanner

@MarkHenderson You're just not interpreting her sentence right. She did it. She is the whole cause of it looking ugly. She must be in 4th grade!

Actually some of our users tell me some pretty insulting things without realising that I'm often the person responsible for the thing that they're insulting

Just as well I stopped giving a shit in about '03

I had a guy do some work at my place a few years ago who used our software. He had no idea of my relationship with the software and hte company he worked for, and he bitched it out to me (which is pretty poor, complaining abotu your tools to your customers).

To which I replied with "Wait, so you don't have to go into the office any more, you're doing your jobs more efficiently, you don't have to hand-write anything, you don't have to copy and paste photos into word documents and juggle digital cameras and a laptop. But you're really, really annoyed by the fact that you have to tap three times to collect a digital signature?"

"Sound like you have it pretty sweet actually" and he just grumbled

Fuck this, I'm going home.

Guess what she sent me?

People have used up their quota of stupid for the day

@MarkHenderson The PDF may technically have part of the font embedded in it ...

hello

@MarkHenderson Also, that's the font she wanted? really? eww.

@freiheit I chose not to comment

I honestly think this woman is being deliberately obtuse because she fears that our software is going to replace her

And the irony is that because of her un-willingness to co-operate, it probably will replace her

But if she just worked with the system instead of against it, she could find herself in a much easier job

Sanity is too much to ask for

Fuck Cisco.

@MichaelHampton Because profits!