« first day (1040 days earlier)      last day (4227 days later) » 

15:16
0
Q: How did SSH get the low port number of 22?

John SmithThe SSH protocol is fairly new. How did it manage to get such a low numbered port? And what happened to the resource which used that port?

keeeeel
@NathanC BANG
@Travis Always wrap your ethernet cable around fluorescent tubes.
It makes the signal better.
Ty. I threw in the RFC because that answers the question, but it didn't deserve an actual answer to begin with
@NathanC it deserves LMGTFY:How are internet port numbers assigned?
HOW IS PROT MADE?!
@voretaq7 1) press butan 2) receive bacon
15:25
@RyanRies Your account has been banned. Ban reason: Because.
Ban will be lifted: When an appropriate bribe is received.
TIL Windows 2000 has no VSS. >_>
Only chased my tail for six months before finding that problem...
1. I though it did. 2. Who still uses Win2k?!
Could be worse:
17
Q: Virtualized MS-DOS computers cant communicate over IPX

DeanI converted 2 MS-Dos 6.22 machines to virtual machines. Both computers have software that communicates over NETBEUI and IPX (I came to that conclusion after inspecting autoexec.bat - It loads IPX stack). I followed this Vmware Forums post and installed the DOS amd_pcnet driver, configured it, an...

This isn't what it sounds like, but...
I need to deobfuscate PHP that's been obfuscated by Ioncube.
I'm trying to tie duosecurity into a web app
15:39
@RyanRies I know people who still use WfW3.11 but they have a 'lifetime' supply of hardware
@Iain ^ WfW 3.11
I remember the computer taking a minute to load Win3.1 - It's strange to go from "Power On" to "Program Manager" in ~8 seconds (much of which is the "bios")
I wonder if surfing the web with an archaic 16-bit operating system is ironically safe because none of today's malware knows how to target such an old system.
@RyanRies I think it's "Safer" from the fact that Javascript in IE 4 was almost non-existent.
@RyanRies: not very hard to test ;p
You constantly get errors about scripts being "too large to run"
I'm sure most modern viruses couldn't run in 16-bit Windows
15:43
@ChrisS DO YOU KNOW HOW LONG IT WOULD TAKE TO EVEN DOWNLOAD A 1MB JAVASCRIPT ON MY DIAL-UP CONNECTION?! :-)
@ChrisS FWIW, I can't find a clear cut answer
@Tanner: I used to when I had a PIII for a second computer ;p
Ahh yes, waiting an hour for that Tiffany Amber Thiessen picture to load...
Telegraph me those titties, internet!
@voretaq7 I was running the above with a 10Mb vNIC, Win3.11 didn't support 100Mb mode of that card. So my Internet connection was roughly twice as fast as the VM could handle.
@RyanRies THE TELEGRAPH IS DEAD. INDIA KILLED IT.
:P
15:45
THE TELEGRAPH IS DEAD. STOP.
Thinking about Win 3.1 makes me want to play Chip's Challenge
@voretaq7 much better
| service_name = VSS | service_description= Volume Shadow Copy | included_with = Windows XP with Server Pack 2 and newer versions of Windows | also_available_for = Windows 2000, Windows XP RTM/SP1 | related_components = Backup and Restore, File History }} Shadow Copy (also known as Volume Snapshot Service, Volume Shadow Copy Service or VSS), is a technology included in Microsoft Windows that allows taking manual or automatic backup copies or snapshots of data, even if it has a lock, on a specific volume at a specific point in time over regular intervals. It is implemented as a ...
Screw it I'll ask on main. Already have Win 2k and Win NT tagged questions anyway...
15:51
@ChrisS I was thinking more /golfclap
but that works too.
16:03
Wow. This bright star managed to create a volume group from multiple physical volumes… where the multiple physical volumes are actually different paths to the same LUN.

It's not working very well.
@MikeyB all of my lol
OK, if you didn't know this existed, GNU parallel is super:
gist.github.com/Supermathie/e3ba316b8be101371938 (i.e. run ALL THESE JOBS but only #CPUs+4 at the same time)
('sem' -> 'parallel --sem')
I seyz:
You want to replace the filter line in your lvm.conf with this: filter = [ "a|/dev/cciss/.*|", "a|/dev/mpath.*|", "a|/dev/mapper/.*|", "r/.*/" ]

He seyz:
Should I keep the following default or replace it? filter = [ "a/.*/" ]

…

THAT'S THE WHOLE POINT OF THIS EXERCISE!!!
@MikeyB I hate that. When you say something and then they ask you about it like you never said it. Because, either they weren't listening (which is bad enough), or they didn't understand and aren't saying anything (so much worse in my book).
@ChrisS Firmly in the doesn't understand anything territory, unfortunately. I kinda almost feel bad for him - I explain and explain but he just doesn't get it.
16:17
@MikeyB lock him up far away from the servers
@voretaq7 He's now the only sysadmin for his group and part of his responsibility is these servers.
WTF HE HAS 100 HOURS OF RH TRAINING??!?!?!?
facepalm
hmmm....
-1
Q: NoMachine NX 4 beta 1 on Fedora 16

AdrianThe latest beta of NX 4 server crashes upon any attempt to log in, with the following segfault: [401227.068201] nxnode.bin[16572]: segfault at 2c ip 00007f1f54d9d0c0 sp 00007f1f23ffec10 error 4 in libnxdimi.so running on Fedora 16, x86 64 bit Unfortunately I can't use 3.5 because I need the ...

should be closed...but not sure how to say "contact the beta team with your bug"
guess the "too localized" is gone
You're running beta on beta, dude. — MikeyB 1 min ago
Two betas, one CPU.
OK that's a stretch.
BUT WHO HERE DOESN'T LIKE A STRETCH AMIRITE??
9 mins ago, by voretaq7
@MikeyB lock him up far away from the servers
@MikeyB I heard you like Betas...
@voretaq7 … at a customer. DON'T ASK.
16:27
@MikeyB why, God? Why?
0
A: RDS, RDWeb, and RemoteApp: How to use public certificate for launching apps on session host?

MDMarraI am pretty sure that you need a cert valid for both the internal and external namespace. If you have .local internally (yuck) then you may have a hard time acquiring a cert with .local in either the Subject or Subject Alternative Name fields. If you can't get a cert from a public provider for th...

^ @RyanRies
Spot check plz
So he's their Ted?
@voretaq7 kind of. Not really.
@MDMarra I believe you are correct.
From client... We need a 4-CPU server! 48GB memory. 2x 300GB disk. We want 4 CPUs with ~2.8-3.2GHz and a 10gb connection for each server. Two servers total. We are going to be doing load testing on these and want them for 12 months
16:35
@MDMarra Ha ha - ".local" bites again
@ewwhite "blah blah, reasonably sane but big config... blah blah ... only want them for a year, at which point they'll be nearly valueless anyways, but hoping you'll only charge us 1/5-1/3 of their actual cost and eat the rest." ??
@freiheit and 2.8-3.2GHz 4-socket systems don't really exist
@ewwhite Overclock them. ;)
@ewwhite Well, sure, they need to be talked into a reasonable CPU choice...
@ewwhite Though, I am seeing DL585 with a 2.8Ghz CPU option, but I'm not convinced it's any faster than a DL580 with 2.4Ghz CPUs...
(for the non-HP nerds, 585 is a 580 with AMD CPUs)
@freiheit Yeah, there's a DL560 Gen8 with 2.9GHz... but only one.
and AMD is never a consideration
16:54
@ewwhite Don't like AMD?
@Jacob I've never seen AMD servers taken seriously in a datacenter for some reason. When I was more involved with server hardware, they just didn't have the scalability that intel had, and you'd get more VMs per dollar with an intel IBM blade than with an AMD one.
@Basil Interesting. We only have Intel gear, and I only have Intel stuff personally, but I've never avoided AMD
I guess that's why the stock is worthless
@Jacob I went to one of their sales things once and they had me wanting to believe, but the numbers never panned out.
@Basil Chopper uses AMD processor for DB servers of some kind or another.
@ewwhite a Xeon E5-4617 runs at 2.9Ghz
@ChrisS I do, they're nice, cheapo, Oracle DB blades - 4 x 16c/512/2x300GB/FC-mezz=£5.5k/$7.somethingK
17:05
@Chopper3 Can you hear me in TS?
@Jacob muted, on a call
@Chopper3 Oh, sorry
been on a really long dull call all day as one of my guys updates all our vCenters - yawn
@Jacob no problem, muted as I say
@Chopper3 What kind of DB work? A bit of everything?
@ChrisS various really, video entitlements normally
17:10
We're eval'ing DuoSecurity, and one of vendors wants an outrageous amount for their plugin and it's really frustrating.
YubiKey?
Better vendor?
No, it's a web app that we use for billing/support
We're most likely going to get the Duo tokens
Anyone using Comcast business internet
@Travis You mean ComCrap?
17:13
@Jacob Oh man! Comcast is going to be so cheesed
@JoelESalas huh
I've been using Comcrap Business for years.
@JoelESalas aren't they already pretty cheesy?
@Travis Yep
@Travis yes
morning everyone
17:16
I've been on Comcrap since they bought ATTBI since ATT bought @Home... So maybe '99 or thereabouts.
Fuck!
@ewwhite You're not having a good week, are you?
@84104 Just about to say the same thing
@ChrisS Holy crap, @Home, that brings back memories!
I had 38Mb service with @Home - AT&T cut me down to ~6 IIRC. I'm now back up to 25Mb... Still not what I had ~14-15 years ago.
17:21
@ChrisS you had 38MB 14-15 years ago?!?!?
@MattBear Well, 43Mbps down, 31Mbps up... but for whatever reason we though it was "38" at the time. Yes
I remember when Napster hit. I was constantly kicking people with dialup off my client.
@JoelESalas Wooo! I'm a huge Monster Cat fan. =)
@dawud Not what I was expecting. I like surprises. =)
@dawud Got all those noted in Spotify. Looks like the week is taken care of!
@DennisKaarsemaker And also handling chat flags. Most importantly.
@WesleyDavid @Rchern Doesn't work for stack anymore???
@Jacob Hasn't for a while.
@ChrisS damn, how much was that connection?
17:27
@WesleyDavid WTFBBQ???
Not the only one who went AWOL as well.
@WesleyDavid hell yeah
@Jacob I can't imagine working with a bunch of asocial nerds in a community manager role is very satisfying
@JoelESalas I wouldn't know
This is a case of thin-provisioning gone wrong
I fucked up
again
17:34
0
Q: iSeries Apache SMTP error

MaKRI'm trying to send mail using either the built-in PHP mail() function or through the phpmailer library, which of course relies on mail(). phpmailer says it "cannot instantiate the mail function" so I'm troubleshooting directly through mail(). I do have mail() working correctly elsewhere, but th...

> I've never heard of RCPT
@ewwhite Not your week, man.
...dear god why?
@ewwhite noooooooo
@voretaq7 As in, RCPT TO
@WesleyDavid yes, that would be my general assumption when given in the context of an email error message too
but apparently That Guy has never heard of RCPT
17:35
@voretaq7 As in put your computer back in its box, give it back to Best Buy, and go make sandwiches at Sprouts?
@voretaq7 As in I can count to potato but my compiler can't!
One potato, Two potato, Three potato, Four!
Ferment, Distill, It's vodka now! Drink 'til you hit the floor!
@MattBear I don't recall at the moment... But I'm pretty sure in the $70-$90/mo range.
@ChrisS wow... thats not bad for now
I ran a MacGuyver-style ZFS overprovisioned/thin-provisioned system out of space
With XFS and a cherry on top!
17:37
@ewwhite WOOOOO
It was a brave new world, @Home was profitable, and they just weren't concerned with squeezing every last penny of profit out of customers... It was awesome.
I always wondered how you were supposed to avoid running out of space when thin provisioning. Now I know!
2
@MattBear BWAHAHAHAHAHA
3
Q: When using thin-provisioning with ZFS, how do you make sure you don't run out of physical disk space?

Mike SprossForgive me if this seems like a fundamental question, but I couldn't really find anything concrete on Google, and I'm not a system administrator by trade. We are setting up a SAN at our office using NexentaStor with an 8-disk RAID Z3 configuration (8 x 1.36 TB drives) and are in the process of c...

@MichaelHampton By rapidly provisioning new disk as you grow.
:)
17:39
in The Bridge, 19 secs ago, by jeffreylin_
@Wipqozn Half of reddit is filled with idiots. The other half is from Stack Exchange.
3
Except I'm running out of space... on the physical disks...
@ewwhite Moar disks!!
@tombull89 Wow.... the rare cogent argument.
@ewwhite thin provisioning is a mix of "pretty nifty" and "fucking terrifying".
@WesleyDavid OK, if this is not surprising enough for your voluptuous life, nothing can be
17:46
@MikeyB If things grow too fast...
@dawud Ooooo...
Lack of stimulating questions on SF lately, it seems.
I guess I done answered 'em all.
but in this case, the client's local guy storage vmotioned data from a thin-provisioned volume to a non-thin-provisinoned volume.
and it expanded all of the VMDKs
17:47
@ewwhite Ohhh, that's bad. =(
and well... teh suck!
So, explosions happened
so I stuffed it back into its original location
@ewwhite Do you graph/alert on rate of change of size of thin-provisioned volumes?
but it wasn't a good process.
17:48
This just sounds like it will turn out as a new "I no longer have AD" question
1
Q: Adding second Windows Server to domain

Force FlowI have a Windows Server 2008 R2 Server (Server A) running as a domain controller with Active Directory, Group Policy, and DNS services. I set up a second Windows Server 2008 R2 server (Server B), added it to the domain, and installed the Active Directory, Group Policy, and DNS Roles. If I take ...

@MikeyB for this... it's a pair of 400GB SSDs...
and a VMDK on top.
Within the VMDK is a ZFS pool
with a ZFS zvol...
and the ZFS zvol has XFS on top of it.
so 400GB physical space. 275GB ZFS pool. 800GB sparse zvol.
@ewwhite So I heard your like drive formats...
vol1/pprovol  refcompressratio      3.12x                  -
vol1/pprovol  written               245G                   -
What's the point of overprovisioning anyway...
vol1/pprovol  used                  245G                   -
vol1/pprovol  available             24.7G                  -
vol1/pprovol  referenced            245G                   -
vol1/pprovol  compressratio         3.12x                  -
highly compressible data.
17:50
@NathanC @ewwhite is a storage god. He probably should be working for the NSA.
[root@Not-pr0n ~]# df -h
Filesystem            Size  Used Avail Use% Mounted on
/dev/sda2             9.9G  2.4G  7.0G  26% /
tmpfs                  12G     0   12G   0% /dev/shm
/dev/sda1              97M   59M   34M  64% /boot
/dev/sda7             2.0G   68M  1.9G   4% /tmp
/dev/sda3             9.9G  4.5G  4.9G  48% /usr
/dev/sda6             6.0G  994M  4.7G  18% /var
/dev/zd0p1            800G  519G  282G  65% /ppro
vol1                   25G  128K   25G   1% /vol1
vol2                  151G     0  151G   0% /vol2
@MilesErickson See that's funny because they have to store all of our data.
lol @ the hostname @ewwhite
so to the software, 275GB looks like 519GB.
@TheCleaner That's actually a good reference for me since I get to install our second DC when the new server comes in.
17:52
@ewwhite dude, storageception much?
[root@y0-mama ~]# zpool list
NAME   SIZE  ALLOC   FREE    CAP  DEDUP  HEALTH  ALTROOT
vol1   274G   245G  29.0G    89%  1.00x  ONLINE  -
vol2   298G   143G   155G    47%  1.00x  ONLINE  -
@ewwhite I suddenly had a flashback to when Stacker, DoubleDisk, et al. were all the rage.
diskdoubler...
Hey, storage is expensive.
@MichaelHampton $3k per SSD
Using two... may need to go to 4
17:53
You got them down to $3K on those cards?
I have to buy my SSDs in 8 packs
@MichaelHampton Oh, these are enterprise SAS SSDs in an HP box
@NathanC of what not to do? Seriously, adding a 2nd DC is too easy. Static IP/DNS, dcpromo, add DNS role if necessary, add DHCP role if necessary, done.
for the FusionIO 1.2TB cards, I got them down to $5300. I bought 4.
see it's getting ready to go bad, quickly.
Should I remove and reinstall the active directory services role? — Force Flow 11 mins ago
17:54
Damn. If only I had $5300 laying around...
Holy shite are you ever getting raped on your SSD prices.
"S3700 400GB SATA 1.8" MLC Enterprise SSD for IBM System x" → $1.4K LIST and we sell for $1K-ish
@MikeyB I don't pay $3k for them.
Oh thank goodness.
(I get them for $400 on the black market)
@MichaelHampton You want one... HP has 29 1.2TB cards left... heading to the scrap yard.
@TheCleaner Yeah, exactly.
17:57
@ewwhite I think I want the address of the scrap yard.
And the Intel SSDs are even cheaper (half that). You can't afford to not buy SSDs nowadays.
I'm guessing closing a DC VM won't work like I want it to eh? :p
*cloning
JoeQwerty isn't too happy with the current devops....lol
@NathanC Do it
Thanks for your response. The reason for the change is that the website will be hosted in a different provider (godaddy) where the domain currently is. — Felipe 23 mins ago
OK, these questions about moving name servers and email because a web site is being moved are a bit ridiculous at this point. 1. If you don't know what you're doing then find someone who does. 2. If you're moving only the web site then change the DNS record(s) for only the web site and LEAVE EVERYTHING ELSE ALONE. — joeqwerty 7 mins ago
18:00
@TheCleaner Go Joe!
Hello
@Manishearth So our little flag spammer seems to have fallen silent, which is good I guess?
@TheCleaner If I want to break the network... :P
Breaking networks is perfectly acceptable behavior. Ask most consultants.
Meh, it'll just be easier to install the roles and dcpromo it
18:04
For server asset tracking, (I looked at RackTables - but no real easy way for me, personally, to import my info.) - what is everyone using? Just a spreadsheet?
@Cole I use spiceworks/spreadsheets
@NathanC Cloning a DC will break all the things. Do not do that.
Spiceworks shows me the basic info and count, spreadsheets for the detailed (like, what's on this one?)
@Cole Yeah, I was researching it and immediately found it to be a horrible idea.
@Cole The accounting department. So far it's worked horribly.
@Cole For asset discovery, Zabbix. For tracking, /dev/null.
18:06
@Cole vsphere client :)
Darn. Guess I'll have to stick to the spreadsheet.
@TheCleaner How many VMs do you have?
I'm trying to organize 80ish locations of servers and what not. Spreadsheet is kinda sucking.
But I'm thinking, I'm kinda stuck with it.
@Cole 4 - I changed to a startup in the spring.
We have Solarwinds Sever & App Monitor - but it's currently running on a single physical box, and adding all our servers to it = kill it.
@TheCleaner yeah I wish it was as simple as going into the vSphere Client - but we have a fair amount of servers + all different owners.
I looked into this as well: sivann.gr/software/itdb
I'm still considering options for our tracking. Granted a spreadsheet would work since our current "inventory" is an access database...
I've used various things in the past. Spiceworks, MS AIS, spreadsheet, Groundworks, etc. etc. Honestly it's whatever tool you'll actually use and not let the info get stale.
Heck, even psinfo
For a "enterprise" level company, I think we need more than a spreadsheet...since no one updates it. We have no idea where our servers are, who owns them, if we have space available, etc.
Maybe I'm just overthinking it.
@Cole uhm.... wow.... yeah...
@Cole this looks nice, I'd like to try it tumblr.github.io/collins/index.html
I'm going to assume I'm trying to fix something that I'm probably just wasting my time with.
@Cole I'm not sure you really want to do this... it would almost be like trying to inventory the town dump.
18:15
@dawud I think that's way too much for what I need.
@Cole you need OpenNMS + an inventory service
@MichaelHampton I suppose, but I have a spreadsheet to start.
@JoelESalas No idea what is OpenNMS
@Cole - horse before cart. Don't look for a tool until IT develops a policy on who's going to update, when/how often, what info to collect/maintain, where it's stored, etc. Then find a tool that fits the need.
@Cole @ewwhite has a huge boner for it, I'm not a huge fan
@TheCleaner well, SysAdmin is supposed to update it whenever we deploy a new server
18:18
I think the way to go is to teach your servers to feed the CMDB with that info, instead of expecting someone to do it manually. I don't know to what extent that can be accomplished
I think @MichaelHampton is right - this is a waste of my time.
but the info is already there: facter, zabbix LLD, IPA, you name it
@JoelESalas OpenNMS looks like a monitoring system ...well, I guess that's obvious seeing NMS = network monitoring system
Yeah we already have monitoring, Solarwinds SAM.
@dawud Most of that's fetchable via SNMP, too. There's things like physical location info that will inherently need to be done manually, though.
18:19
@NathanC A+ 10/10
We don't do automation here.
3
@freiheit not if you have the proper custom facts in place
I've asked, but been told no.
@dawud Ok, sure, but at some point some human will have to have told some software, somehow, what the physical location of the box is. You can push that into all sorts of places.
@Cole are they from the past?
18:21
@dawud long story, but short answer: yes.
@dawud (Personally I prefer that it be in the destination DB, because I don't want to lose track of which rack my box is in just because I reinstalled it)
@Cole Give up while you're ahead. If they don't have one now and have been functioning, might as well keep it that way.
@freiheit for me, that is done in provision time, through a cobbler snippet, so no manual intervention
It's like... trying to build a car without an assembly line.
Yeah, blah - oh well
18:22
@freiheit (the location fact, that is)
@Cole Do they have oversized eyebrow ridges? Pronounced occipital ridge? How far from the past are they?
I apologize for even bringing this up, again.
@dawud Ok, I can see having that info be there, too...
@MichaelHampton So, like, building some sort of crazy high-end $300k+ car that the company makes at most 4 a year?
@MichaelHampton Artisanal handcrafted servers!
Heh.
When we ordered the server yesterday, I pointed out that I could've gotten the same processors for $95/each rather than the ~$250/each the company was selling them for when added to the package.
Guess what they went with...
@NathanC where did you order the server from?
Damn, I'm in the wrong business.
@NathanC Never heard of that site
Why not from Dell directly?
@Cole Me either.
Uh, money I guess.
Well it's an R710 - that's an older model eh?
A new server from Dell was probably double that
But hey, I don't do the purchasing or the budget ...so...I work with it :P
18:36
Who was talking about my boner? Was it @joelesalas ?
6
18:47
Yes, I have a hard-on for HP, OpenNMS, ZFS, VMware (not vCloud) and CentOS... Anyone got a problem with that?
I wish it was 4:30
I've had a productive afternoon. And by productive, I mean I just spent most of the afternoon reading notalwaysright.com
2
Oops.
@Felipe - No offense intended but this site is for "professional" level questions as stated in the FAQ. It is not intended for DNS 101 questions. There's nothing wrong with you asking questions and there's nothing wrong with people answering your questions but you need to be open to the "push back" that a basic question like this is going to get. A question like this, that could be answered by almost anyone with a basic understanding of DNS, "clutters" up this site and drives more advanced professional level people away because they're tired of sifting through these basic questions. — joeqwerty 37 mins ago
Has it already been discussed in Meta that the new close options don't have a "this is not a professional tech question"?
Or do we simply choose "home setting" for anything that is "ask your IT admin" or "you obviously need to hire a pro"
easiest answer for a while :) serverfault.com/q/524141/1435
Too many yesses.
18:55
If it were 2009 that would get 74 upvotes @Chopper3

« first day (1040 days earlier)      last day (4227 days later) »