« first day (1311 days earlier)      last day (3956 days later) » 

Bob
Bob
12:00
Ah. Just love it when half the updates randomly fail.
12:19
Morning gents
12:57
It's a beautiful morning in Boston.
serious wind storms in MI
Nice temp, pretty much perfect
but it's been 70, and is going to be in the 30s tomorrow, so one of those warm front meets coldfront deals.....
0
Q: Software/System aspects of disaster recovery and business continuity on a VPS

user2346536I'm a developper who just decided to run its next web app on a VPS instead of a complitely managed server. So I'll learn. When at my workplace, I can focus on developpement. Whatever happens, even if the entire building is the target of a nuke, System/database/network admins and production teams...

man... This sounds like "explain your job to me, kthxbye!"
...System/database/network admins and production teams, will magically resume normal business state...., It's not magic, it's hardworking system admins doing their job :D — BigHomie 10 secs ago
I think I've seen that before... but WHERE
Bob
Bob
@Jacob HMMM (just in case you're actually wondering: xkcd.com/1354)
13:13
@Bob I'd rather see the original on wiki. What time is it for you? You should sleep more :P
Bob
Bob
@Jacob Time of the xkcd one? Probably last Friday.
@Jacob 11:15 PM. Not too bad.
But I'll probably be up till 3.
@Bob I was up until 4:30 and now I need to fill out MAH tax organizer for my account. Hopefully I don't screw it up too bad.
Bob
Bob
Ahhhh taxes :P
I have a couple months before those, thankfully.
@Bob I have until tomorrow
Bob
Bob
Ouch!
13:18
@Bob I didn't get the paperwork until a few days ago. I've been busy. Worst case I file an extension, as I don't owe any money. I just won't get all the taxes I paid back.
13:29
Working with uni network admin to forward some ports
So slow.
One email every 2 days.
Who decided to name this exploit Heartbleed anyway? It sounds so sinister, why not HeartAttack? Is it cuz it's bleeding information from the TLS Server?
@BigHomie The people at Codenomicon
Or Heart Arrhythmia? Which is an abnormal Heart beat right?
A rumor has been started that I am deliberately trying to sabotage a system... other than settings all event logs to archive, settings the size of the security log to 2.6 GB (for max entries before archive) what else can I set? The system runs off a file share so how about auditing on the folder?
@BigHomie Because it "bleeds" info?
13:33
@Jacob Are you asking or telling?
@BigHomie Both?
@Jacob Well, by bleeding I mean that it's slowly leaking vital information via this exploit (CWIDT), so that could be one reason for the name
@Travis I think disproving the "rumor" and figuring out how it started would be best
@NathanC I know who started it. It is harassment now actually because she constantly questions my "day off" even after I worked a 40+ hour week. I can disprove it simply by event logs (I wasn't logged in at the time)...
and you can't clear individual logs. It's an all or nothing deal.
@Travis Go to HR
Yeah, and it also logs that you cleared said logs when you do it. :P
13:39
howdy folks, my topic was closed, but I have absolutely no idea why: serverfault.com/questions/585645/…
@NathanC Yeah, they don't get that...
@NathanC Guess I'll talk to my immediate supervisor and tell him whats up. then next will be a harassment complaint to HR
I am a professional "system admin"
the question is on topic
@Travis Good idea.
so, is it possible to reopen the question?
@mbrownnyc Post on the meta, please. Mods will look at it.
13:40
@NathanC okay thanks
Unless people vote to reopen, only a mod can unlock it on their own.
@Travis I started that rumor.
@BigHomie gasp
@NathanC It needs to be edited for reopen voting to be enabled
14:01
@JohnD The real gasp is finding out how long Codenomicon kept that little gem to themselves, and trawled the 'net with it, especially since blackholes like facebook and others are switching to https
14:15
I'm installing Fedora in a VM to learn systemd, but damn it's so slow! At 1G of RAM, it shouldn't be that bad...
@KevinSoviero dynamic vhd much ;) ??
@BigHomie Maybe...
Even still, it's REALLY slow. Like wait-over-5-min-to-load-empathy slow
@KevinSoviero moar SSD
@KevinSoviero Ouch
Eww... I don't think I like systemd. Here is the command to boot into runlevel 3: ln -sf /lib/systemd/system/multi-user.target /etc/systemd/system/default.target Oh ya, that just rolls light off the tounge.
*tongue
14:30
@KevinSoviero up arrow to edit
@NathanC Spoke to my immediate and the company owner. Owner just laughed and said don't worry about it. He knows it's B.S. and the person it came from is even more B.S.
@KevinSoviero systemctl enable multi-user.target is a little nicer.
@Travis That's a good sign.
@MikeyB Does that change the default? And if so, someone needs to update the recommendation in /etc/inittab
@KevinSoviero It's Fedora. YMMV.
14:40
Don't let Chicago become Detroit!!! - Weekend Violence Leaves Four Dead, 37 Wounded
@MikeyB He's pretty laid back
@BigHomie Or is Detroit really not that violent?
@MikeyB But I also wanted it on record that I complained since the CFO told me we needed to intentionally sabotage the system a few weeks ago...
@Travis Good, good. :P
@Travis This coffee. It smells like sabotage.
So… our LOVELY customer has decided they'd rather go with OEM Office 2013 instead of VL (against our recommendation) because "it's cheaper."
That means each PC we image for them now takes an extra 45 minutes to register it to the specific user account, download the dedicated installer SPECIFICALLY FOR THAT KEY and install it.
Thus we've upped our quote for them for new PCs by $112.50 per PC for the extra time involved.
Good move, guys.
14:44
I want to get my RHCSA and RHCE... I wonder how my employer will feel if I stop working to study...
@MikeyB More $$ for you!
This is an https page, right?
"stop working to study" wat?
@KevinSoviero no
@KevinSoviero Erm, no
I was joking
Bob
Bob
14:47
@BigHomie It is sinister.
@MikeyB "dedicated installer SPECIFICALLY FOR THAT KEY" HAHAHAHAahaha
you poor bastard
Bob
Bob
Sooo... I finally got Symantec Endpoint Protection Manager working again. After a full reinstall (!!).
Feels like I just made a deal with the devil :\
@JohnD Find out what's different in the installer and set up a script to modify it? :P
Oh wait, more money.
Urk. Downloading gigabytes of installers would absolutely kill the DSL (!) connection here.
@Bob true, I'm sure there's just an INI file or something in there with the key in it.
It's thundering like crazy here! I love it!
does anybody know what is this http://xiabox.com/ site? Apparently it mirrors SF:
https://serverfault.com/questions/588889/zabbix-server-mysql-cant-allocate-shared-memory-access-denied http://xiabox.com/questions/588889/zabbix-server-mysql-cant-allocate-shared-memory-access-denied
15:00
@MikeyB wSounds like you guys didn't do the best job selling the right solution
@MikeyB How many PCs a month are we talking?
I'll admit up front I don't deal on the workstation side that much...but for the XP to 7/8 migrations are your IT depts. using Easy Transfer or USMT? I'm going to be assisting another company with their XP to 7/8 and wanted to know what I'm getting into.
15:17
oh man! We are trying to move a domain to office365.
We wanted to prepare the existing e-mails accounts there so we verified the domain to be ours with a TXT record.
And from that point on we can't send e-mails anymore to that domain, because office365 routes them internally and not to the current MX record. ARGH!
@faker Based on the Dilbert cartoon above, I give you 6 minutes to complete the task.
@ewwhite We were pushing VL and were tempted to tell them "if you don't use VL we won't do it at all".
THEY DINNA LISTEN
@faker well what did you expect? (true, at least google handles that better)
@MikeyB smack them, then!
Does anyone really understand Linux filesystem ACLs?
I sure don't.
1
Q: Read-only files, prevent deletion

LOLapaloozaHere's my situation. I'm running a WordPress website that uses w3 total cache. A large portion of this website is at first dynamic, and then static. So I'm caching these pages as static HTML in a folder, let's call it /home/test.com/wp-content/cache/page_enhanced/static/. So each file in that...

@ewwhite Yep. All over that.
@ewwhite I hate people who put web server stuff under /home. I'm a FHS nazi.
15:24
/var/wwwwwwwwww ?
nope, `/srv/www` :P
@ewwhite /var/www/playboy.com or /srv/www/brazzers.guru yeah
@MikeyB that it verifies that the domain is owned by us but still send mails to the server specified in the MX
how are we supposed to cutover accounts if we can't prepare them?
and I'm also curious if only our domains (==same account) or all office365 accounts now route it internally
@faker They don't have a documented migration process for this?
@faker :) why are you using the O365 mailboxes already then? Just use your existing mailboxes on Exchange/whatever. Meaning only O365 mail should be routing "internally on O365"...so don't use it yet.
15:30
@dawud I usually do like /var/www/vhosts/example.com just to make it cleaner
I made it even more organized by doing /var/www/vhosts/example.com/pages so I'd have a dedicated log directory for each domain...and a folder to put the SSL keys
@MikeyB none that I could quickly find but I'm also not running that project. I just get asked why we can't send mails to $domain anymore
@NathanC not in a SELinux enabled machine, right? Ubuntu or the like?
@dawud Oh please that's easy enough to fix under selinux.
@MikeyB I don't like to put unneeded extra stuff semanage fcontext -a -t http_conf_whatever
@dawud No, I don't usually use SELinux
15:34
@dawud You need to change your perspective, that's part of the installation.
@TheCleaner we don't, we haven't even created a single mailbox yet, just added the verification TXT record and it started routing internally
It's been a while since I've really touched webservers and the like...mostly just to get familiar with them.
@MikeyB there are already defined paths for certificates, logs, keytabs, etc in the policy. It is easy indeed to modify it to fit some customer's demands, but I rather not.
@dawud well you just have a different structure then that matches existing expectations? No different than what @NathanC is doing.
@MikeyB In my experience, it's good in the long term if you keep things as close as possible to the refpolicy. @NathanC dir structure is completely fine, it just does not work for me.
15:39
I've honestly never liked SELinux to begin with. Sure, disabling it is taboo in the IT world, but...getting it out the way instead of debugging for hours dealing with so many commands and their associated - args (in typical *nix fashion) is more appealing in some cases :P
@ewwhite It most certainly is. Me and a coworker were talking and wondering if it was this bad all over (like Chicago) and we just don't hear about it b/c we live around here, so all we hear about is the sad Detroit stuff
Funny how Rev. Al never shows up for these kinds of things
@NathanC Yeah, it's worth spending the time to get familiar with it, only takes a few hours and it's not that bad. Just unfamiliar and new. There's some good helper tools to change policies to fix your requirements as well.
@BigHomie Truth... Although Rev AL. has his own issues... (Mob connections)
apparently we are just stupid and should have done this o365guides.appriver.com/m/8585/l/…
@ewwhite Didn't know that, but he's already been pegged by my generation as all political anyway
15:42
@BigHomie Yeah, nuts... just came out - npr.org/2014/04/08/300617881/…
@ChrisS Mmmm...frise...
ouch
>


1. Code blocks are for code. Please use quote syntax for quotes. 2. Markdown provides a nice system for numbered lists. Use it. 3. Sentences should start with capital letters. This includes your second list item, above. 4. Markdown provides a mechanism for emphasis. You can use that instead of slashes. 5. The name Stack Overflow is spelled thus, with capital letters and a space.
@Iain what's that about?
It's a comment (not mine) upon someone's question
15:53
@faker routing internally how though? External domains to yours should still be using your MX records, even if they are hosted on O365. Your own domain mailboxes on your O365 account will try to route to O365 hosted mailboxes, but if you didn't create them yet I'm not sure how there is even an issue. Your mailboxes on Exchange should route email to your domain locally on the Exchange org. Still not sure how you are experieincing what you are.
Is it possible to buy these power cables from any stores? NEMA 5-15P to IEC 320 C19
This hotel's bathtub fills from the ceiling... how odd.
@DanilaLadner yeah… ca.startech.com/Cables/Computer-Power/External/… $20+shipping
@DanilaLadner um, whatcha doin?
well from brick and mortar stores.
well, i have IEC 320 C19 to IEC 320 C20
15:54
@DanilaLadner Sameer has them.
bigger HP servers used to have 5-15->C19
well i am pretty sure he does, i am in NY5 need them like in a few hours -)))
@DanilaLadner Ask Equinix
@DanilaLadner Shirley some place like Fry's has them?
@MikeyB nope
@DanilaLadner seriously, ask the data center. This is what they're paid to do.
Haha, i did. they said since I do not have a direct account with them but through Pico, they cannot sell them to me -)))
15:56
Then you ask the tenant to approve the order.
Come on, man.
I could arrange Logicworks to help you.
yeah, i've tried, support said would get back to me in a few hours, i just thought i could get them myself in a few hours.
Why is select * from blah; considered bad?
@KevinSoviero For the same reason that Windows is considered bad.
@ewwhite Pretty positive we will be buying 4 servers from you ewwhite, btw
15:58
@MikeyB It's bad for political reasons?
@KevinSoviero No, bad in specific situations, not in general.
@MikeyB he's in a data center nowhere near a Fry's.
@DanilaLadner show me the money.
@ewwhite Oh NY5 is a datacenter?
@MikeyB Equinix, Secaucus.
@ewwhite I was all like "oh he's doing this":
15:59
@DanilaLadner The Logicworks people at at NY4.
@ewwhite yeah it is next door.
They likely have the cables.
@KevinSoviero Because it's horribly inefficient?
Let me talk to Pico again, and then if not I will ask your guys.
but how did you get into this situation? What servers do you have that need a C19/C20?
Blades... 4-socket machines?
and your DC isn't on 208V?
16:00
@KevinSoviero It's considered "bad" because you should be requesting only the columns you actually need from the DBMS.
@ewwhite blades
@DanilaLadner Seriously? Plugged into 5-15?
Fucking ancient datacenters man.
yeah -))), well it is actually a quite new dc, that rack is not production, no 220V there.
i have for 208V cables but they are c19 to c20
@KevinSoviero Say column_1 and column_2 in table blah has been indexed - SELECT column_1,column_2 FROM blah; can be satisfied by scanning just the indices, but SELECT * FROM blah; still requires you to scan and read the entire table (some of which will probably need to be read from disk and into memory)
16:08
k, pico support dude will bring in 40 minutes, he is coming to ny4, thanks all. But is it really impossible to buy something like this in electric supply store?
@DanilaLadner Graybar.
but you have to plan, man...
I order custom cables or right-sized cables for DC work.
So Tripp-lite and Cables2Go...
@ewwhite yeah, i did not expect apc to have 120V
and can get them overnight.
@ewwhite I should of asked them.
@MathiasR.Jessen And if you happend to actually need all the columns?
16:10
@KevinSoviero No perf bash, but you still risk having the columns returned in a different order from what you expected
@MathiasR.Jessen k
Something that's making me rather uneasy...been a week without email antivirus and we're still running without on-access scanning within forefront
(apparently it's corrupting our ERP system, but...okay...)
I can't prove it, but I swear that some of the apps I've downloaded on my phone have sold my number to telemarketers. I get calls from various 800#s daily, sometimes 3x a day from the same #. I block them on Verizon's site, but still it makes me wonder HOW they get the number.
@TheCleaner Ah, if only you could deny permissions to an application like you can on BB.
@TheCleaner Random dialers
16:24
@MathiasR.Jessen Yeah, the real failure there is assuming you already know the order of the columns when parsing the result.
@MathiasR.Jessen Most languages allow an "associative" return, meaning that the SQL driver will return the column names and you can reference them in the code (for example, running mysql_fetch_assoc() in PHP would let me use $foo["column"]
@MikeyB Well, there are rooted apps that will allow you remove permissions an app has AFTER install. But I just have to wonder the amount of data collected and distributed by free apps even if they are ad-supported. I know I'd even be willing to collect and sell the data an app I built collected if I knew it was legal and would make me $$$.
@TheCleaner It's not legal unless the end-user knows this data can be collected/sold
@MikeyB @NathanC I know that, but I've also seen actual code - running in production - exhibiting absolute disregard for common sense and sensible practices.
@NathanC who reads a EULA?
@voretaq7 - see this?
Is this true??? "They should train their flight attendants. The doors open inwards. Because of cabin pressure (higher then outside pressure), that door has thousands of pounds of pressure holding it in place. Your average human can only pull with 20-40 pounds of force. Not even Schwarzenegger on his best day could open it. "
16:35
@TheCleaner He has a point.
@TheCleaner It's probably more of the "scaring the passengers" thing. I for one wouldn't be comfortable seeing someone yank and pull at the door at 35k feet even though I know they can't open it.
I didn't know you can't open a door in flight because of that inward thing...but I can assure you that man wouldn't have been able to see out of his two black eyes
and I mean that i would have beat his skull in...not black eyes 'cause he's black.
right, external mails still reach it.
But mails sent from domain a.com ($corp o365 account) to b.com (now also on $corp account, verified but external MX records) now bounce.
o365 says that's a domain here but mailbox does not exist, bounce! It does not forward it to the mailservers configured in MX records by default
@NathanC Good to know. Next time I wish to exit an aircraft I'll shoot through the window.
16:45
Client: The phone system is down!! HALP!
Me: Call your phone vendor.
@ewwhite Correct answer.
@Travis Yay decompression!
Client: The phone vendor says that the phones are down because the Windows Server isn't licensed.
@ewwhite snerk
@ewwhite Answer: "Then license it, dumbass."
16:45
Client: This was the message that Gus got this morning. Once he hit activate the system started working.
Oh dear...
facepalm
@ewwhite Windows 7?
Me: Nope... I just added the license to the server
Although, I had 2012 R2 whine at me the other day because it didn't want to activate
16:47
The phone vendor is trying to say that hitting "activate now" made the phone system work
I went through the automated voice thing and it happily accepted the key, but still didn't activate it according to slmgr
@ewwhite Is it a windows based phone system?
@ewwhite Was it to a point where Windows was like "screw you, no logon for you until you pay up"?
@Travis It's Shoretel... so there are hardware phone units.
the Windows Server is just a management interface
XP used to do this and it was rather irritating...
16:51
Well, that sure looks like XP, or something from the same era.
2008 R2
@ewwhite Interesting
Fucking hacks.
Local IT people and vendors are a pain in my ass sometimes.
@MichaelHampton That's non-aero NT6.
can you use a # in a freebsd password? It seems the ! is not liked very well
16:57
People turn that off? Do they like living in the past?
I have software that doesn't like it
I turn off Aero. Don't like it...
I turn off Windows. Don't like it...
haha
@Travis Dude, I've put the fucking escape character (yes, <ESC>, Ctrl-[, 0x1b) into passwords for people.
17:02
@MikeyB ! doesn't work though :)
17:12
I know I ask basic Windows questions, but @mdmarra won't answer my calls anymore.
3
When running updates on some Windows 2008 R2 systems this week, I noticed that some updates stalled or simply could not complete without a hard reset.
How can I control that or make the update process more deterministic?
@ewwhite Nope, you're at M$'s mercy
Although, I've never heard of updates that required a full shutdown and cold boot
@NathanC I had two instances of updates this week that stuck for over two hours.
so me being an IT professional, I just turned the VMs off and back on again
@ewwhite Maybe it's a specific one?
@NathanC how the hell would I know?
Event Viewer?
17:22
I mean, in Red Hat, it's like, "yum update stuff"
@ewwhite The windows update screen doesn't tell you which one it's on?
@NathanC sometimes this happens when I'm logged-off and at a "Configuring Updates" screen
I had that happen today. One update took 30 minutes to install. Rest went quick
so how do Microsoft people in big environments handle this?
@ewwhite Ah...okay. Sometimes the event viewer will give you which one it's stuck on (look for the install success messages from MsiInstaller)
Other times...not a whole lot you can do :P
17:29
@ewwhite Isn't MSCM or something like that always the answer?
@84104 You mean SCUM or SCCCCCM, or SLURM or...WSISSY or STUSSY?
@ewwhite WUSSY (the Windows Update Server and Slipstream Yoinker)
Just now, two new laptops with Win7. Booted into Windows after updates, they sat at a black screen with the spinning circle cursor. Left like that for a while, no go. Connected with Bomgar, tried to reboot, no go.
Hard power off, hard power on, all OK.
@Travis errr yes it does (9.1-RELEASE-p4) Maybe you have a derpy application that can't handle that, but that's not FreeBSD's problem.
@MikeyB Not with this crap I'm using apparently (nas4free)...refused to join the domain until I made it a passwd without a!
@Travis There you go - a derpy application.
17:35
@84104 tasty!
Or the other SLURM
Meh
@Travis Nas4Free?
Hmm... you can do better
@ewwhite How about FreeNAS?
Tempted really just to load up ubuntu or something on it instead ...
How do you guys load balance DNS requests
@JoelESalas With a Citrix NetScaler
17:38
@Travis facepalm or FreeBSD?
We have two servers but one server is always first in resolv.conf and so it gets bombarded
Spent too much time trying to get it just to work ...and all I want is a simple CIFS file share so I can put back ups on it.
@JoelESalas with multiple NS records.
@freiheit lol wat
@MikeyB we have two NS records for our internal zone but it's still slamming one server
@MikeyB Could put BSD on it too... I was just more excited about having a cool little GUI
17:39
preferentially
use anycast for dns
@MikeyB But their support just flat sucks..and their documentation is worse than Microsoft
@Travis WTF you talking about? NS records are well documented.
@JoelESalas We use a Load Balancer to load balance DNS. The resolv.conf/DHCP conf lists the load-balancer VIP IP first, then a standalone DNS server in a different subnet, then an off-campus DNS server we have an arrangement with.
@MikeyB NAS4Free
17:40
@freiheit I don't know how I feel about that. I want the clients to choose.
@JoelESalas Are you always handing out 'ns1, ns2' to dhcp clients? Most (all?) OSes will try the nameservers in listed order.
@MikeyB Yes that's how we're doing it. I was thinking we should have puppet randomize which one comes first...
@Travis So try FreeBSD or FreeNAS. Haven't used FreeNAS myself but heard good things about it.
unless there's a smarter way.
@JoelESalas If you want the clients to choose, then hand the IPs out in a different order to the clients.
17:41
@MikeyB NAS4Free are the same creators of FreeNAS before they sold it to a company. Reviews on the N4F were better than FreeNAS...
@Travis No FreeNAS
@JoelESalas Pseudo-Randomize.
@JoelESalas rotate in resolv.conf. Not sure about Windows.
@ewwhite Can you be even more cryptic with your response...offer an alternative...
@JoelESalas um... timeout and rotate
17:42
Anycast is perfect for dns. hand out one address, they both get a chance to respond.
@JoelESalas "one server is always first in resolv.conf" ← there you go, you've even identified the problem
@Travis Depends on what you need.
Place to store backups
@ewwhite 8 TB H/W RAID 1+0
NexentaStor, Napp-It, QuantaStor.
probably the latter.
Are they free?
17:44
but I think that's a 4TB limit
all have free tiers
In smb.conf:
[backup]
path=/storage/backup/windows/%m
M-M-M-M-Magic.
@Travis Just install FreeBSD and samba and DIY. It's really not hard and you can earn a few neckbeard points.
@Travis He speaks in riddles
@MikeyB That's just a function of our CM, we can change that but I like the rotate method
@NathanC nas4free.org
@Travis Never used it, but I've used FreeNAS before. It's pretty slick and it works well
About 5-6 years ago we had a "test" one set up on an old P3. It'd peg on 100mbit with LAN transfers (this was before gigabit was widespread to non-enterprise)
Plus, it's completely free...it doesn't have built-in RAID but you'd use a real controller for that anyway
17:55
This one has software RAID but I'm using the controller
Oh, and it was already mentioned...meh
simple samba would be all you need
Yep. This does that...plus allows RSYNC and all kinds of stuff

« first day (1311 days earlier)      last day (3956 days later) »