« first day (1271 days earlier)      last day (3996 days later) » 

00:00
@ewwhite Yeah but we also have snakes that eat crocodiles to keep us safe
@MarkHenderson squirrrrm.
yesterday, by Mark Henderson
user image
@MarkHenderson that makes me itch all over...
@MarkHenderson needs more photoshop
@ewwhite There's a video too if that would make you feel better
00:05
@MarkHenderson send it!
@MarkHenderson Yeah that's pretty sick
Who else uses Vagrant
My bad; the video was a slideshow
I thought it was an actual video
@JoelESalas never heard of it 'til now, but does it work w/ HyperV?
00:09
@MDMoore313 Allegedly yes!
@MDMoore313 Apparently it does now
@Andrew rofl
I saw that you have an excellent reputation on serverfault/stackexchange and it just so happens that my colleague reached out to you before as well. If you're interested, I can give you a call today between 7:00 PM to 8:00 PM CST or tomorrow at any time after 12:00 PM CST.
Shameless.
@MarkHenderson @JoelESalas that's awesome, someone should update their Wikipedia
00:11
@ewwhite Square?
@JoelESalas LinkedIn.
@MDMoore313 HyperV is not my thing
@ewwhite linkedin itself?
awesome
@JoelESalas no, not awesome.
17 mins ago, by Zoredache
... If you're open to considering other options at the moment, I'd like to set up a time to go over the position that I had in mind for you. I support LinkedIn's Site Reliability Engineering team and I thought your experience with maintaining Linux servers ...
Brotha-man is spamming ServerFault people.
@ewwhite Should I be insulted?
@JoelESalas That's okay, HyperV isn't bad
00:14
@JoelESalas yes
Anonymous
Anonymous
"Why the IT Industry sucks - Part 1 - Archie's System Administrator experience " an aussie's story
@PatoSáinz I wish he wasn't try eat his mic. The mic goes below your mouth, not directly in front of it...
@Zoredache Depends on the kind of mic
Anonymous
@Zoredache i agree lol
Anonymous
@PatoSáinz I love that fucking video
Anonymous
@JoelESalas sadist
Just discovered an interesting DoS against iOS. Visit one URL, device is locked up tight as a d-rum
drums, of course, are easy to get in to in general, and not even sealed to external air.
Anonymous
@RyJones 0day?
@PatoSáinz I'm gonna say 0 hour, I just found it a bit ago and confirmed it on a second device a bit ago
Anonymous
00:37
@RyJones does Apple have a responsible disclosure programme?
@PatoSáinz unclear to me, but my employer paid me to find it (incidentally) so it's their work product
Anonymous
@RyJones oh damn it!
Anonymous
You could have made a lot of moneyzzz
Anonymous
@RyJones does it work against latest iOS? iOS 7.* only or what?
It works on iOS7 latest and iOS6 latest.
I'm not digging up an iOS 5 device to test on
Anonymous
00:39

 The DMZ

A serious place where infosec is discussed PS we don't do hard...
eh, I'm good
looks dead.com in there
Anonymous
@RyJones is the bug something like "goto LOCKTHEDEVICECUZWEAREAPPLE"?
Anonymous
I wonder if mobile safari is still vulnerable to the IRC XPS attack
I hit it, I repro'd it, I checked an older version, I'm done digging as I have work to do
last time I reported a bug it took Microsoft months to say "not our problem"
actually, Microsoft's security team said something like "don't use malware like that"
4
Anonymous
@RyJones lol
Anonymous
00:53
@RyJones could you please post your findings (not the details) of your microsoft and apple shits on /g/? I'm sure it will wind up a nice shitfest
Anonymous
also, XP is almost dead ;_;
@RyJones FWIW, they don't treat all bugs like that, don't give up hope
@PatoSáinz XP is almost dead from an MS standpoint, It's been dead from this sysadmins standpoint.
@MDMoore313 since that time, I have developed a much closer and more respectful relationship with Microsoft :)
oh, that page trims my smartassness. marc.info/?l=ntbugtraq&m=102090979805785&w=2 has the original PS: fuck you
01:07
God, I fucking hate Java.
@cole you and me both
Fedora + Java + Mozilla = fuck me in the ass with a cactus
@cole you might sell tickets to that
I get 64bit java working and it's like "lol Juniper Network Connect needs 32 bit java libraries!"
@RyJones we'll hopefully it'll pay for the medical bills after my brain aneurysm
I'm fighting a Java certificate problem at the end of a bunch of rubber bands
the machine with the issue is a VM attached to Jenkins talking to a VM with gerrit
all of which I control by the Jenkins web interface
01:15
Your life sucks more than mine
that was not the argument I was making, but I will agree in order to be agreeable
I would much rather be dicking with Java on my laptop
than that
she called him over, the proper response was to leave
Wow
@RyJones I love how Java's response to security is to make it refuse to run anything that's not signed with an expensive certificate unless you jump through hoops, and even then only if it's got an attribute manifest
It is such a pain because nobody has updated their apps
Anonymous
@MarkHenderson Set to medium security. Laugh at the fact that they're trying to install Mcafee now instead of the ask toolbar. Hope for Oracle's eventual collapse to be soon, and move on with your day.
@MarkHenderson from a security perspective, turning the whole thing into a whitelist greatly reduces the attack surface Its not a particularly stupid idea, but its definitely not the complete answer or an excuse to stop figuring out other security bugs (no idea if they actually do this)
As long as the damn thing still prompts before running anything, I'm not worried about it. The cert signing thing is good for people who can't be trusted to not just click "ok" on every dialog that pops up.
@MatthewIfe DO you use libvirt to manage your containers?
Ok Juniper - I have the 32bit libraries you say you need installed
fucking. work.
01:29
@ewwhite yes.
@ewwhite well, libvirt for kvm. Some containers.
@ewwhite my current designs avoid using containers in their 'complete' form. Just partial ones using a manual combination of cgroups and namespces.
I typically dont need a full range of services to run what I need. I'm using it to segregate/delegate the provision of CGI backends.
I did a bunch of LXC work a 12-18 months ago... and now I'm reading RHEL's knowledge base. They're suggesting using virsh - virsh -c lxc:///
@ewwhite yeah, thats fine for most use cases. It might not give you the complete option set of lxc though.
Well, I know the lxc-* commands. I think the original tutorial I followed back in the day was: Virtual machine stacking Using LXC on top of ESX
To give you some idea what I'm doing, its (in effect, not completely) unshare -n /path/to/some/runscript.sh and that invokes a service thats listening on a IP brought up in that namespace.
SCGI normally goes into that service.
The way i'm doing it is totally custom because EL6.4 doesnt export the namespaces used in processes which broke the traditional model of doing it.
so you're just using it for specific processes and isolation
01:34
Yes. The runscripts peform cgroup manipulation.
how long do they live?
My main factor for using a network namespace is to make it completely impossible for a user to bind to an address on the host which is meant to be used for something else.
Well, the user has some control over it but typically they live for the runtime of the host.
Its significantly easier doing what I did in 6.5
It's interesting to see the different approaches to this. wiki.centos.org/HowTos/LXC-on-CentOS6
and Ubuntu has its own way...
but this plagues Linux technologies... a multitude of ways to do the same thing... No standard.
IF you are intending on just launching a single service, or just a couple -- the unshare command may be something to look into.
@MatthewIfe this really needs to be its own environment.
Postgres + java + Pentaho + some web portal stuff on its own filesystem structure.
01:39
Yeah.. in that case probably nicer to go with something pre-made for this.
This probably wont matter in you're case, but if you can manage to readonly bind mount most of the root filesystem where all the libraries live, it costs nothing memory wise to load.
let me look at one of my older ones...
How the shit do I install Java 32bit alongside Java 64bit -errrgh
@MatthewIfe I was using OpenVZ templates: openvz.org/Download/template/precreated
untarred into the chroot.
@PatoSáinz Yeah, it's not ready for use yet
then creating the LXC config file...
lxc.utsname = web
lxc.tty = 4
lxc.network.type = veth
lxc.network.flags = up
lxc.network.link = br0
lxc.network.name = eth0
lxc.network.mtu = 1500
lxc.network.ipv4 = 172.20.3.29/24
lxc.network.hwaddr = FE:92:40:71:90:53
lxc.rootfs = /srv/lxc/web
lxc.mount = /etc/lxc/web.fstab
lxc.cgroup.devices.deny = a
# /dev/null and zero
lxc.cgroup.devices.allow = c 1:3 rwm
lxc.cgroup.devices.allow = c 1:5 rwm
# consoles
lxc.cgroup.devices.allow = c 5:1 rwm
lxc.cgroup.devices.allow = c 5:0 rwm
lxc.cgroup.devices.allow = c 4:0 rwm
01:44
Yeah. Allows for a lot of customization doing that.
Unfortunately you lose the benefit you get from shared memory mappings between the namespaces.
Its still way cheaper than a VM for memory allocation though. Especially if you're using (the default) optimistic memory allocation strategy.
So even containers have a variety of use cases.
Of course ;). Its important to actually understand a 'container' is nothing more than different namespaces & cgroups. That way you realize sometimes its plausible to pick and choose what parts you want and what parts you dont.
interesting stuff... but I still believe in standards! If someone on SF asked, "how do I build an LXC container on my CentOS 6.x box?", they'd get a few different answers
For joe-user I would suggest LXC nearly all the time.
"chroot on steroids"
I don't think cgroups were introduced to the average user well.
again, I'm RHEL-centric
01:48
I'm really hoping for more in the way of user namespaces before I truly commit to LXC. Plus theres no such thing as a 'security' namespace yet which can be unfortunate for SELInux.
LXC is not secure. Dont give users root and expect to have a good time. User namespaces are a means to offer some security (although the namespace subsystem in this case has some way to go to prove it really is as secure as it intends to be)
In a user namespace you can give someone root and it still be secure (they are root in their namespace, but not the parent one). Its actually pretty interesting to see it working :).
what's the ideal application of this tech?
Of LXC?
sure
It doesnt replace openVZ. Its a nice way to provide managed parallel environments.
right, the hosting folks will still do OpenVZ.
01:52
When user namespaces becomes a thing, its in a much better position to be akin to openvz/solaris zones.
-1
Q: domain-search value on my clients computer

birgirI need help with configure my DNS and DHCP server - I have spent some time searching on google but have not got any informations regarding my problem. I have an Ubuntu 12.04 server running and use BIND as the DNS server. In my client Ubuntu 13.10 computer I get my domain in the search value se...

lol
@ewwhite also then a lot of what people are doing with full virtualization may become more moot. Especially where there are large numbers of VMs being utilized with practically the same environment.
ZFS/Btrfs tricks could also be used to provide different writable snapshots between each instance, so a nice way to manage storage efficiently.
@MatthewIfe yes, ZFS ;)
(should) also give you the memoryo benefit too doing that.
oh, interesting... yes.
01:55
3
A: domain-search value on my clients computer

Craig WatsonThis is not what search in resolv.conf does. It specified the DNS servers that your PC will use for DNS lookups, and the search domain used for hostname-only lookups. Example: nameserver 192.168.50.11 nameserver 192.168.50.12 search mynet.local In the above snippet, the servers 192.168.50.11...

@FalconMomot should we edit it to wesley.bike?
@MatthewIfe I've never used unshare... interesting.
any other examples?
If the whole user namespace thing works out, you'd be able to take a esxi with 128GB of memory potentially handling at least an order of magnitude more 'vms', possibly two or even three (cpu permitting)
sure.
So, run unshare -p -n
In your new shell remount proc (as it still is giving out data of the parent namespace)
@Andrew if it might help him find one, please.
Now running ps -A gives you a whole new list of pids (htis is how lxc does this bit).
As for networking, thats a little more complex but still nothing terribly hard.
do I need to unshare a command?
unshare: invalid option -- 'p'
01:58
Yeah, bash is probably OK. Shoudl do this as root for the example.
me and my old-ass server
Net should work.
In the parent namespace. Run ip link add type veth
Will produce two devices veth0 and veth1 (probably).
yes, two
Get the pid of your unshared bash. Then run ip link set netns <pidofbash> dev veth1
So is there a better replacement for ps -aux?
02:00
k
If you check your parent namespace, ip link ls shows only one veth, veth0. If you check the child namespace, you see veth0 now in there. These pairs are linked like a bi-directional pipe.
right
So you should up veth0/veth1, and just assign an IP into veth0.
This is how a container communicates with the outside world.
Note: Parent namespace linux effectively no longer recognizes packets coming from thie child namespace as coming from itself. It uses the 'forward' chain in the host to manage it.
In fact, the child namespace actually has its own iptables chain too.
interesting... this is also what I see on my existing containers.
14: veth4vVoOW: <BROADCAST,MULTICAST,UP,LOWER_UP> mtu 1500 qdisc pfifo_fast state UP qlen 1000
    link/ether fe:9c:be:09:16:1f brd ff:ff:ff:ff:ff:ff
(done the old way)
So, it may be necessary to either enable packet forwarding in the parent namespace, or creating a bridth and adding veth1 to it.
@ewwhite veth pairs were patched in by lxc years ago.
This isnt how kvm does it though.
02:03
> the easier and non-racy version
wtf, wikipedia editors...
This is (basicallyi) what i'm doing in my namespace stuff.
@MatthewIfe aren't you frustrated at the pace of development?
Because now the user inside of that namespace can bind to say '0.0.0.0:80' and it doesnt 'consume' that value in the parent namespace.
@ewwhite its been there years ;). I'm more frustrated by the user namespaces (frankly its a lot of work I suspect) but that patch has been floating around I believe since 2009 before it finally got accepted last year.
@ewwhite with what you're doing, keep a close eye on user namespaces, systemd and LXC integration with it all. Because you're likely going to be able to get a lot of work through it.
I've not really thought of a highly available mechanism for all this yet -- although thats also (slowly) coming.
Theres a trickle of patches coming in slowly to the kernel to offer support for suspending processes and resuming them on different hosts. (this is partly what the 'freezer' cgroup is intended for)
ah yes.
Its broken and incomplete as it stands.
Yet, if that works out you'll be able to regularly snapshot a whole process tree, and providing you've got the same underlying data exported elsewhere in theory it would be possible to resume from teh snapshot kind of like a migration.
Also useful to know, if you're smart with your exported volumes (assuming mostly everyone requires the same libs in some environments) you can just mount the one tree and export it readonly to multiple hosts and multiple instances.
Then you can update from one 'writable' host and signal all other hosts to drop pagecache to refresh.
In either case (I dont know your business) but if you're competing with other people to deliver on price, you'd be in a position to make a significant profit for yourself, still undercutting traditional vm deployment competition AND beating them on performance.
kce
kce
02:12
@ewwhite - Paging our favorite vSphere expert...
http://serverfault.com/questions/579888/vsphere-5-1-u1-client-the-command-has-timed-out-as-the-remote-server-is-takin
Seen that one before?
I think a lot of startups will lap it up.
@MatthewIfe possibly... Familiar with Docker?
@kce I spent as little time on 5.1 as I could.
Yep. Its LXC with a package management system.
@kce So I'm on 5.5 and have the improved SSO features
kce
kce
@ewwhite - Yeah. They released 5.5 during the middle of our upgrade to 5.1U1 and then we hit our "feature freeze" period. We should be free to change stuff in April.
02:15
@kce this is support ticket material
let me read it again
@kce funny thing, these problems made me move to the Linux-based vCenter
The support for Windows vCenter installations is dwindling
kce
kce
@ewwhite - The VCA? Our NetApp tools require Windows unfortunately.
kce
kce
@ewwhite - wow. so I can connect using PowerCLI.
I guess the rest of our staff needs to learn powerShell now.
Haha. Man. That's awesome.
OK. Now it works.
Methinks that Office 2013 is going to kill a lot of the dodgy "edit PDFs!" crapware around.
kce
kce
What the hell.
02:24
:(
kce
kce
Whatever.
Hmm. I wonder.
@kce Working now after the reboot?
kce
kce
Oh. I think I figured it. I'll come back tomorrow and post the Answer if you jerkwads upvote my question. :D
@kce Heh, already did!
kce
kce
@ShaneMadden - Lots of departments run their own AD and don't use the central one like our department. The vSphere SSO had picked up all their domains and was using them as identity sources. As a matter of house-cleaning, I removed them and finished writing the question. At that point the web client had finished removing them and then poof things magically started working.
I can't prove it exactly but that's the only thing that changed in those 15 minutes or so.
(at least in our infrastructure)
02:34
@kce Ahh, so it was trying against each? Any default domains set up in SSO?
kce
kce
Nope!
Hence the LookupService log with all the AD entires.
@kce Gross.
kce
kce
@ShaneMadden - GUBERMENT!
@kce Oh, I know the feeling, used to work for a county gov't.
kce
kce
Later dudes. Beer o' clock.
@Andrew Yup, more terrible from goto.
I don't get why people haven't stopped using goto. Are developers this bad?
If OpenSSL turns out to have one of these as well, then the tinfoil hat explanation of the NSA planting these bugs will be looking pretty plausible.
if it was an anonymous "here is an open source library, enjoy" contribution, maybe.
seems more like "works good enough, why test more?" though.
so it seems like a combination of 1. authors messed up and 2. nobody check these things
03:11
@Andrew They do now after Apple dropped the ball.
 
4 hours later…
06:58
@ScottPack The use of goto was appropriate... the excessively idiomatic code is what got them.
07:22
morning
@dawud night
G'day
07:48
Hail, Lord Darkstar!
08:14
I'm just glad that open source code is peer reviewed, or else a vulnerability like that one might have been in play for nearly 10 years...
@RobM peer review schmoo
I just find it amusing. Or I would if it wasn't so serious.
peer review of open source is erm yeah
I remember "peer review FTW!" as one of the many reasons why open source was magically better than closed source back when i was involved with the borg and had linux advocates telling me how that made me the devil on a daily basis
which is why I find it amusing now.
08:33
yeah, nobody has time to actually review that code, and if they do, chances are they are going to want something (not necessarily monetary) from it.
@FalconMomot: they ought to hire a bunch of OCD folks of the right sort to code review, and nothing else ;p
At the very least, it would be nice if e.g. RedHat would have a bunch of nerds do quality assurance on crypto stuff and everything else that is included in a very basic install, given that they are after government contracts (and not just US government, where NSA backdoors might be considered a feature that just makes operations smoother all around... ;))
Now this is the story all about how
Our life got flipped, turned upside down
And we'd like to take a minute, so just sit right there
And we’ll tell you how all how we moved to VMWare.

In Xenserver Enterprise born and raised
In the server room where we spent most of our days
Chilling out, maxing, relaxing all cool
And all shooting all the servers into the pool
When a couple of updates, they were up to no good
Started making trouble in our neighbourhood
We had numerous crashes and the users got scared
5
@JourneymanGeek who will do the salary-paying?
@FalconMomot: maybe run it as a security consultancy? ;p
08:49
with no clients?
I never said it was a good idea.
;p
Dan
Dan
09:20
@LucasKauffman This is the most wonderful thing I've ever read
09:30
@Dan Clearly you haven't read much :p
Dan
Dan
@MarkHenderson I've just spent too much time with XenServer
It spoke to me, man
@Dan I had to work with a XenServer/DRDB setup once. It achieved a mind-boggling 80% uptime
Dan
Dan
@MarkHenderson :D
Our shared customer was so pleased that they gave us $100,000 to build a system for them
09:33
80%?
And the guy with the Xen service went out of business as that was his last profitable client
@tombull89 OK fine it might have been like 85%
Dan
Dan
@tombull89 In XenServer terms that's like 7 nines
Oh HyperV sighs
09:34
But it took five hours to fail over to the other node
I don't know if he was just jerking off hoping the server would boot OK or what
But once a week it would be down for about 3-4 hours
during business hours mind you. Not maintenance windows.
Anyway, anyone here know AnyConnect licensing? On 15.3(3) AnyConnect works just fine, but on 15.4(1) it is telling me it's not licensed and is locking me out of it. So I've rolled back to 15.3(3)
Is it a bug in 15.3 that I can AnyConnect without any extra licenses?
< Me> | I'm doing my monthly system report again, how's memcache doing these days?
< Dev> | much fine, many gets
I hate doge.
8
It'll get out of fashion soon enough.
or at least, thats what you'll tell yourself if you like sleeping at night.
 
1 hour later…
10:56
@ewwhite I use libvirt's variant of LXC, which isn't entirely compatible with "regular" LXC command line tools, but seems to be easier to work with overall. For this I can make my own templates, e.g.: yum --releasever=6 --installroot=/var/lib/libvirt/filesystems/centos6_template groupinstall Base
 
2 hours later…
13:06
Morning
Dan
Dan
'sup
It's Wednesday
...........and that's about it.
Dan
Dan
Heh
Just got myelf a season ticket so, for the first time ever, I'll be a rail commuter
I let vanity get the better of me and I went for First Class, too - I'm justifying it because I can have breakfast and somewhere to work
Student had "opened" a doc from Hotmail that she'd sent to herself - saved it and then quit. Course that saves it to the local machine so did an epic bit of digging, pulled it out of her Tempoarary Internet Files and dropped it in her work area like a bawwwws.
I've done that a few times with outlook
13:13
@Dan ohhh fancy.
Dan
Dan
@cole Well, if I'm going to work my ass off contracting I may aswell get some benefits
When I was looking at jobs in the city, I was looking at a Commuter rail pass (it's the train that takes you from the 'burbs of Boston directly into it) - My apartment is a 5 minute walk (literally, train runs behind my apartment) to the stop.
So I looked up the monthly cost of a commuter rail pass $350+
ouch
That doesn't include the cost of what it would be if I had to transfer to another train/bus after I got off at South Station in Boston
So I was like nope already have a car payment.
Dan
Dan
@cole Yeah....you don't know want to know what I just paid
13:15
@Dan hopefully you're making more to justify it?
Dan
Dan
@cole Put it this way, I've paid for it already this week ;)
To be honest, I think I could write it off at the end of the year/find a company that would pay for half/whole thing.
@Dan Glad to hear it :)
Would be kinda nice to take the commuter rail into work and not have to drive.
Dan
Dan
@cole Only got another 6 weeks on this one, though - so here's hoping I can keep it up!
Wouldn't need to drive the station/pay for parking as it's at the end of my street.
/me wishes he could quit his job now
Dan
Dan
@cole Yeah, I don't mind driving but this drive is a fucking bitch and the traffic is just stupid. I've only done it 3 days and I can already feel the journey stressing me
13:17
@Dan yeah Boston is a PITA to drive in, which is why I don't work there.
oh ok Windows VM - just freeze - that's cool
Didn't want to do work today or anything.
Dan
Dan
@cole Kick the shit out of it
/drop kicks laptop
Shaved my mustache off last night, this morning wake up with a pimple right there. fml
Dan
Dan
:D
afternoon
@cole JustShavedProblems
13:23
@NathanC pretty much
So, decommissioning AD CS just to re-create it...yay pointless work!
I was stupid to think using sha512 for the CA cert would work with anything
other than browsers that is
We use SHA-1 for most shit
@cole Yeah...the only way to change the encryption method is to nuke from orbit and start over
@NathanC indeed
13:29
@ewwhite merge that with flightradar24.com and you'd have quite something.
@tombull89 exactly.. the company is promoting the HD video imaging as a way to help businesses... but seriously, it's creepy. See: firstimagery.skybox.com/hd-video
Uh...how does it help businesses?
@tombull89 watch the video
Oh my god... the founder of that company is one of my high school classmates.
3
Damn, I went to school with some smart people.
and they have Linux and DevOps positions open.
Every so often, I think I'm doing well... then I look at Facebook.
Kids from my high school are pro athletes, entrepreneurs, researchers in Antarctica, marine biologies (for real)... Then I think about what I've contributed...
13:48
I'm putting in for a vacation day on Monday.
I just love "undocumented features" ...especially when they just happen to resolve a problem that shouldn't have been happening anyway
My whole life is one undocumented feature then.
14:04
morning
hey @Basil
we purchased the Hitachi storage.
14:15
@cole nice :) The HUS-VM?
@Basil not sure yet
Management doesn't tell us anything
yay
I'm going to ask today though
cool :)
You're going to hate raidcom
I have a love hate relationship with it. It's literally infinitely better than the absolute shite management and CLI that came before it (CCI on the USP-V made me hate humanity), but it's undocumented, has useless help files and error codes, only works some of the time, and won't warn you before you do something that would brick a million dollar storage array.
Die die die! ...oh...and DIE!
-2
Q: Is crossover cable still needed?

evachristineDo I still need a crossover cable for connecting two PC's directly? (NIC <-> NIC)? If not then from when? When was the time people walked in to a computer shop, buyed 2 PC's and they could connect them directly without cross cable?

@Basil pants = shat
They're sending us for training
14:38
@TheCleaner What's wrong with that? Other than the shitty English, it strikes me as a perfectly cromulent question.
@HopelessN00b "Your question is especially clear" - isn't?
@ChrisS Yeah, I'll fix that. Thanks.
@HopelessN00b Oh, new word... very nice
@cole That's good. They tend to train people on the management they want us to use, but it's hamstrung
@HopelessN00b look at the 2nd part of the question and the OP's comments.
14:40
@ChrisS [classic] Simpsons word.
@TheCleaner Holy Moly! He/She's an employee of a Major computer company.
@TheCleaner Yeah, well, that part's useless (or related to the fact that crossover is for 10/100Mbit), so it cna be editted.
@ChrisS Apu Nahasapeemapetilon Mike Smith from HP support?
@ChrisS ? That in their profile? I didn't see anything like that. I just laughed at the "ok...2005 it is then" part.
@TheCleaner Their IP addy... which is why I'm not saying which company.
Jumping Jeebus... putting contractions where they don't belong again.
@HopelessN00b No... He'll be calling me again this afternoon though - wanting a ProCurve Chassis back.
@ChrisS ah...I see.
14:48
@TheCleaner Fixed. Now it's a somewhat useful question.
This happens to me about 1 in 10 replacement parts HP sends me. The return label is in a green envelope-thing on the outside of the box - it gets ripped off in transit. So I have to get a replacement label. Ship the box back with new label. Get harassing calls about returning a part that's already been returned (and I see was delivered to them last week).
The real question should have been "Can you still make a crossover cable without looking up the pinouts online?"
@TheCleaner Is it Green and Orange that get flipped?
@TheCleaner Pfft. I can't make any kind of cable without a pinout guide.
@HopelessN00b Did the mods here tell you to be nice now or else? ;)
14:50
@TheCleaner Not to my knowledge, no.
@ChrisS yeah. It's basically A on one end and B on the other. I remember a telco guy in the 90s that would wire everything up A and it confused the piss out of me.
@TheCleaner Same here - previous place I worked was all wired A...
Then we got a new patch panel out back that was only labeled for B...

« first day (1271 days earlier)      last day (3996 days later) »