FrontPage Is Too Damn Localized

yay, one more row of apps FOR TEH PRESIDENT! \o/

oh and btw:dcurt.is/3-point-5-inches

seems someone really did think about how to enlarge the screen without screwing the device.

@Chopper3 Wow, what a world-class fuckass. As soon as I get more votes tomorrow, I'll make sure he goes back to 1 rep.

haha

@Chloe Yes, that makes sense. Broken English(?) but it should make sense to an AD admin. What's the context?

@HopelessN00b It was mine - working on someone elses site and need the Office 2010 GPO's. I rewrote it in the end!

Ah, gotcha. So British English, rather than broken English, probably. :p

Bit of both, I'm not feeling great this morning

@Chloe: monthly friend?

@Chopper3 anything new and interesting you re working on currently?

@JourneymanGeek I think I may be pregnant

@MichaelHampton That guy's not actually asking about FrontPage extensions, he just doesn't know how to read [a damn product version]. You should change your close vote reason to ~ herpderp SO-quality question from developer who can't read.

@Chloe: Oh, just wait till you get the cravings

@HopelessN00b I'm still waiting for the SE devs to add the close reason herpderp stupid developer.

@pfo not really no sorry, mostly DR stuff and some capacity/storage uplifts

did you guys switch to UCSs already?

@Chopper3 Picked up Battlefield 3 (PC) from Tesco last night for £21 so I'll be looking forward to getting on that this evening and the weekend.

I only put those links in my answer for personal amusement. They were most certainly NOT for your own education, so please don't click on them.

@HopelessN00b Excellent - groping for the masses

@tombull89 good news, I'm Chopper3 on that too - I'll let my son hand you your arse

@pfo oh and we are doing new stuff on UCS yes, just not old platforms

@Chopper3 chuckle yeah, I'll probably be standing around gawping at everything.

@tombull89 the premium pack is amazing

i like the groping warning signes in japan: m5.paperblog.com/i/3/32286/…

@Chopper3 Looking forward to Armoured [Whatever It Is]

@Chloe got that, HUGE maps, nice new foliage too

I have a date with the XBox on Friday night when I get back

@Chopper3 What's the AC130 like?

@Chloe not as much fun as you'd think tbh

@Chloe I love Ziba Tower from the Close Quarters pack - fucking nuts, especially with 64 players, and the Gun Master (or whatever it's called) mode is super fun

@Chopper3 Yeah, I was saying on here the other day, I was reluctant to get Premium cos I didn't think I'd enjoy CQ, but it's bloody ace

@Chopper3 How many content packs are there? I'm looking forward to Wake Island and the big one with the mast.

And it's great for going through the challenges, too. Shame they're a bit smaller on PC, really

@tombull89 just get the premium pack, it gets you all of them, much cheaper

Has anyone ever noticed this on a Google image search result page?

I just edited the crap out of this.

Isn't that what I wrote?

@MichaelHampton this guy is an utter moron, check the other question he posted today about not knowing how to add a freaking user account on his box ...

@HopelessN00b oh yes, that's the rule, the more rep the nicer the person, look at me

He tried to enable sudo on a Red Hat box, that's just asking for trouble.

IT Directors should not attempt to sysadmin.

@MichaelHampton Uh, what makes you say he's an IT Director? I mean, anything's possible, but there;s nothing to support that idea, and plenty to suggest otherwise.

this stuff belongs on unix se

That question still should remain closed.

That level of dumbass makes me think he/she/it is probably an HR director.

@HopelessN00b Incompetent, bad attitude, irritating - sounds like an IT Director to me

@Chloe That just says upper management to me.

@HopelessN00b Management, in general

Fuck it, it's just everyone who isn't me!

Nah, my current boss is pretty damn good. So was my last one.

@MichaelHampton he's a photographer

Director my ass. my Google-fu says he's the OWNER.

i wish so badly tim would have release and a retina imac yesterday :/

@Chloe "Fuck everyone who isn't me" ... I like it. Sounds like a girl I want to get to know better! Do I need to buy you a drink first?

@MichaelHampton yep, one-man band

My next comment will be "hire a sysadmin".

@pfo quiet intro early Oct ;)

@MichaelHampton do it

@MichaelHampton Owner of a one-man business run from his parents basement?

just had an retina to play with for a few days

at first i switched to native 2880x1800

but couldn't do it for more than a few hours

    Hodges, Max  [email protected]
    3-12-23-308 Kamimeguro
    Meguroku, Tokyo 153-0051
    JP
    7184049223    Fax: 7184049223

A former IT consultant from Texas, Max is the company's business architect and creative visionary—combining business acuity and technical know-how with a talent for design. - whiterabbitexpress.com/products

hahaha

Hmm. I think I broke the site.

hipster douche.

Looking at Google Street View, it's a house.

Or what passes for one in Tokyo.

"Special order anything from Japan. Get a Quote No-Obligation" - how about we ask him for quotes for "Linux for Dummies' books etc?

@MichaelHampton future matchstick-rubble-pile ;)

@Chopper3 101 uses for a tentacle?

@MichaelHampton Too snarky as a reply?

You really need to consult anyone who knows the first thing about Linux. If this is actually for Systems Administration in a professional capacity, as the FAQ states is the scope of this site, I recommend hiring such a person to add users for you. These are occasionally called "administrators." When they administer a computer system, they are occasionally known as "systems administrators," if that helps point you in the right direction.

@HopelessN00b Maybe just a bit.

@HopelessN00b i'll upvote!

Too bad SF can't have a "general refence" close/flag reason.

It can be deleted by a mod later, I just really think the fuckass in question ought to read that.

me too

Alright, time to head off to work... I suspect there'll still be a Texan hipster in Japan to snigger at derisively when I get there...

going to grab lunch.

is this me being dim or the other guy? serverfault.com/a/427044/1435

The questioner was being dim, but the other answer was better, I think.

@Chopper3 The other guy mostly, but I think you're missing the point a little yourself. He seems to be wanting to make shit magically faster by cramming everything onto a magical RAM disk panacea, on the magical cloud/virtualization panacea.

@MichaelHampton agree PFO's answer's better but you shouldn't be putting swap in a ram disk right?

No, that defeats the whole purpose.

@HopelessN00b I tried to change my name back today, but you can only change once every 30 days. So, I'm stuck like this for a bit :(

his comment - aghh!! - serverfault.com/a/427044/1435

@Chloe Like 'Quantum Leap'

@Chloe Go to SuperUser or any other "non-Chole" profile and mash the "copy profile to all sites" button.

@tombull89 Interesting hack - I'll give it a shot

I'm about to drop below 10K.

You'll loose anything unique to your SF profile, but you'll get your name and gravatar back.

Well, it will be shortly

Proved my point though - got up to 205 profile views since last Monday, from 138 since last October :D

afternoon all

@HopelessN00b wish I could upvote that twice

@Chopper3 same here!

@HopelessN00b good response really, but it's wasted on guy like that - sadly.

@Chopper3 your "ok, you know best" repsonse was pure gold, though. Couldn't have put it better myself.

he's moaning about a downvote, bless

@pfo I'm sure it'll be wasted on him, and frankly couldn't give two shits about him, but for some odd reason, felt an uncharacteristic desire to point [potential] future knowledge-seekers in the right direction.

@HopelessN00b So much for your future as a BOFH.

Uncharacteristic on my part, so probably more related to procrastination and just not wanting to address the clusterfuck of clusterfucks I have at the office in the environment I... can't believe I actually willingly walked into.

And knowing full well it was a clusterfuck too. Gawd, I'm a dumbfuck.

@HopelessN00b who is it you work for. Listening to your woes really makes me want to ensure that I don't use their products at all fo fear of contamination by the stoopitity.

@HopelessN00b he's come back - with more twaddle

@Chopper3 Hell of a day to use up all my [down]votes 4 hours into the day, huh?

not sure I've ever used all my up or down votes

@Iain You around?

Hi everyone?

I have a typical doubt.

Say that apache is getting 500 request at a time.. At that time for some reason my machine (where apache is running) goes down. . It means that 500 requests are lost? Are there any way to prevent this?

is it long enough?

@GroovyUser can you see where it says 'The Comms Room' up there?

This level of st00pid is actually (depressingly, horrifyingly, despairingly, etc., ...) basically very common when you have to interact with "warehouse workers" or user on a plant floor. In every such environment I've ever been, I've been convinced we have the dumbest people in the world working for us, and that we must have to seek out st00pid to achieve this level of "duh" ... and I end up with the same conviction at every new place too.

Plus, the Execs... always asshatted and pennywise, poundfoolish. Everyhwere. 'tis the nature of upper management.

@Chopper3: So this is not the place to ask this kind of questions?

@GroovyUser If there's nothing to recieve the requests I would think they get lost into the digital ether. To stop the requests being lost you'll need failover servers.

@GroovyUser "doubt" makes almost no sense in English. We use the word "Question".

@GroovyUser Yes, and no. Yes, you lose those requests and no, there is any way to prevent that, short of expensive, complex HA/loadbalancing on a cluster.... other than not letting the machine "go down."

@GroovyUser You should ask on the main site

@tombull89 Well, depending on what exactly he means, simple failover wouldn't do it either, as the currently active sessions would die with the machine going down.

@Dan I really don't think that's gonna make a good main site question, somehow.

@HopelessN00b HA doesn't have to be "expensive", something like CARP can detect a downed host within a fraction of a second. You'd still lose some packets, but TCP should be able to cope.

@HopelessN00b True. What I meant is that you'll need some way of "coping" with a downed host.

@HopelessN00b Sessions would have to be stored on some kind of centralized storage.

@HackToHell @ChrisS: I have a failover server and load balancing for it. As @HopelessN00b says, I want those 500 requests not be lost. . Are there ways do so is my question..

Sorry, asking this question on main site will make it too chatty!

@Dan makes this quote a bit more entertaining now:

oh.

@GroovyUser You can make the failover quicker, which would reduce the 500 to 100, or 50, or maybe less.

@ChrisS Yeah, this is what I was/am thinking he's talking about, preserving the 500 sessions active at any given point in time. Not gonna happen for less than mega bucks, I don't think.

Also, the client should detect a lost packet, and if your HA server responds quickly enough the client wouldn't notice. This assumes a well designed failover system.

@ChrisS: So your solution is to save my sessions some where. And use them?

There are questions on the main site already about fail-over and load balancing. However asking "how do I load balance apache" is a pretty poor question and would require a bit of thought, time, and likely money.

@HopelessN00b Put the session information on a DRDB volume, or SAN; NFS would work, though that's another SPoF.

We did something to do achieve this in a 6 nines SaaS..., "hosted networking provider" environment, so it's possible, but I think the system cost millions of dollars, just for that functionality alone.

@GroovyUser You keep HTTP Session information? (not just people connected, but Web Application Session Information)

@tombull89 Mehee

@all : By the way I use amazon aws

@HopelessN00b Isnt that something along the lines of 30 seconds of downtime a year?

@HopelessN00b: Oh that is too expensive!

@HopelessN00b To actually get 6 9s you'd have to do more than what I'm proposing (start breaking out big bucks generally). My proposal would hit ~4 comfortably. Probably get 5 on a regular basis. But it's a lot of careful design.

@tombull89 <=6 seconds of downtime a year, actually.

@tombull89 30s == 6 9s; Oops.. Mess that one up. 5m == 5 9s. 3s == 7 9s.

Yup.

@GroovyUser Why did I get panged ?

@HopelessN00b herp. woah.

@HackToHell Probably did a "H" "Tab" looking for Hopeless

:P

@tombull89 Yeah, and they actually meant 6 nines, because a lot of our SLAs refunded mega-bucks (like millions of dollars to multiple clients) if we fell below the 6 nines metric.

It's going to be one of those days..... Last night I suggested the fix to the new TLS problems were to disable TLS (it was supposed to say disable Compression)... Oy

Is it cool to post a link to my question in here? Not getting any attn atm.

@Switz Bear in mind that this is ServerFault chat, not StackOverflow. How long aga was it posted?

9 hours

No

I'm on a FiOS home network.

Likely to be a home/residential limit.

your server?

@Switz "it can only handle up to ~40 websocket connections at once" <-- Why?

pokes markdown

Through my tests, the dropoff seems to happen at around 40 concurrent users.

@tombull89 I think markdown pokes you!

@Switz Sorry, I didn't ask how you arrived at that number; I asked why.

Well, I assume it's more than capable of handling more, but I don't know what the best way to go about that is, hence the question.

Or how to investigate further as to why.

But I'm open to any suggestions.

Alright; now we're getting somewhere. I suck at JavaScript, so we'll need to find someone who doesn't suck, and find out how to profile your app, find the bottleneck. Then we can figure out what resources to throw at the bottleneck to fix your capacity issue.

I don't believe the app is bottlenecking any more than most node.js apps.

The queries are very basic

So there inherent flaws in node.js that prevent any sort of reasonable scaling?

@Switz Well, if it's not a coding bottleneck, might the low hardware specs to be to blame? What's the OS overhead and such on the system's resources (and its 512MB RAM)?

@ChrisS just found a document online going on about node.js being great for millions of connections

@Chopper3 ;]

I believe there might be inherent flaws on concurrent connections of a single process. I have yet to test multiple processes – so I figured I'd gather myself first and ask around. I've yet to find many resources on the matter. The node.js holds up fine, it's the websockets (socket.io) that fails. Pages still render on the client, but there's no connection as far as I can tell.

Yep, the ram could definitely be an issue, although locally on my much more powerful machine I saw the same results

@Switz I think that realistically your server may not be powerful enough or your ability to test and/or measure the maximum connections may be at fault

What about network contention on your ISP's residential-grade network? You have FiOS, which is fast, but it's still on a pooled (and usually throttled) circuit, and residential broadband is notorious for having problems with lots of concurrent connections for exactly those reasons.

@Chopper3 Yeah, this, exactly this. Even if the hardware's up to the task, a residental-grade network connection probably isn't, not for concurrent connections.

Very possible. But then how, as a college student, can I go about figuring out what I need to launch. Like I said in my post, there will be a lot of hits in the first few weeks, but a lot less soon after. I don't mind spending a bit of money to get a nice server for a month or so, but I want to make sure what I get is capable of handling what I need.

@Switz pay a pro

@Switz As a college student, you say? Most universities have nice, better-than-enterprise-grade connections to the county's backbone.... might hook it up there, if you can get permission or figure out how to do so without asking.

Nope, this is a personal project. I want to know how to do it, so I'm going to do it. It's just a matter of getting there.

@HopelessN00b: not a bad idea. I can ask around, although I know my university is tough with that stuff.

@Switz Nothing wrong with "hiring a pro" for personal projects and learning from what they do. If the project piques my interest, I've been known to work for beer+pizza on stuff I consider "easy" or "enjoyable."

@Switz if you want a site to be production-quality on day-one there's no shortcut, you need experience - the only other option, caused by either inexperience or low budget, is to do what you can and fix as you go - that's a model that almost everyone (facebook included) did, there's no shame in it, as volumes (and revenue/VC) comes in you can make things better

I have a client on the phone now...

I've been programming for years, but I've never scaled a node app before.

@HopelessN00b what's this "country backbone" by the way?

@Switz There's also that adage that it's often better to beg for forgiveness than ask for permission... maybe consider that before asking around too much at your uni. ;)

@HopelessN00b Mostly not till the weekend.

And the client is pushing back on a VMWare setup I've proposed.

"We want to try to do it ourselves... So we understand how it all works."

@Iain In that case, I work for an umbrella company (~$2 billion dollar market cap), on assignment with a particular subsidiary. The umbrella company's stock symbol (which I'll delete shortly) is...

A 3-host VMWare installation with a 80+ user migration to Exchange from Sendmail.

No Active Directory

@Switz You can build a kit-car in your garage but if you want to mass produce them you might need help (testing, productionisation etc.)

@ewwhite oh bless, they don't want to pay you

I understand that, but that doesn't mean I can't try :)

Would you advise a customer to do this on their own?

@Chopper3 Pretty sure it's like "the cloud" a hand way generalization. In this case for "those really fatty pipes mostly owned by the US government connecting various major cities and regions."

@HopelessN00b They're asking me if it is going to damage our relationship.

@ewwhite I think you'd be much better-positioned to know the answer to that than anyone here... or anyone other than "them," even.

@ewwhite depends on their skill level

Running two linux servers I've installed... they are not fully trusting of my recommendation to virtualize.

Still on 10/100 networking.

still have lots of serial devices.

There's a Windows 2003 cluster on HP ProLiant G3 servers and an MSA1000 there

@ewwhite they paying your full rate?

@Chopper3 So I didn't know how to cleanly say, "Installing a new VMWare cluster and establishing Active Directory, migrating mail to Exchange and building a storage solution aren't trivial tasks"

@Chopper3 They're asking me if I'm okay with them trying to figure it out on their own.

@ewwhite You just did

@ewwhite say no

I did... And they're going to think about it...

but the call started because they are trying to build a new website... and needed some newer PHP modules than what are installed on their current web/mail server.

And they haven't made a decision on mail, and we're not in a good place to upgrade from CentOS 4.5...

so they have a deadline of 10/1 to get this new site up...

@ewwhite I'm sure you know this but as a consultant you can never be too honest and communicative with clients, but then you can also never be too delicate either

snicker

and I suggested either buying some decent G7 hardware to get started down the path of virtualization..

or just spinning up a CentOS 6 image in the cloud or in one of my data centers.

I'd agree, G7 is very cheap at the moment - as are the options

I've serviced this place for TEN years...

and I think the two internal IT guys are feeling left behind

and want to try to tackle the new BIG infra project on their own

@Chopper3 You really think? I'd disagree on the delicacy thing. I make a point of not being delicate, and seem to have a rep for being a real honest, stand-up straight-shooter of a guy for it. Almost seems dishonest to let people keep holding that impression.

@ewwhite well leave them to it but let them know you can be called in when needed

I would think that the company would like to hear your honest opinion since they've been paying you for 10 years...

but things have changed since 2002... I don't think the internal IT guys should try these things without having some training...

@HopelessN00b I'm surprisingly tactful in a professional setting, certainly compared to me on SF :)

@ewwhite Consider offering training/teach-as-you-do-it for an inflated rate?

I see the best approach as hiring the expert, and having the experts train you on what was setup.

That's what I offered.

@ewwhite say that then

and I don't think they know what they're getting into.

worst comes to worst just firebomb them

@ewwhite Can't say that enough, IMO. "You really don't know what you're getting into. Let me say it again..."

You think it's okay to say that?

I mean, there's a reason these people pay me...

Well, you've just seen my take on delicacy and tact as a consultant, and how I approach that, so with that in mind, yeah, I would definitely be repetitive on that point. And be sure to know I'm offering them the best of both worlds at inflated rate, but definitely repeat that point.

and I think that in a non-complex single-site company, it's difficult for internal IT to get the experience needed to be able to hammer something like this out.

is that a bad assumption on my part?

"look you want the training, but you also want it done right... lemme do both for you for a little bit more money, and you'll get the best of both worlds."

I'd be saying "I hope you know how much I care about you and your business, I know what you're trying to do and support skilling-up your staff as they do this work but I want you to know that I'm here for you if you need me as it would be very easy for even the smarted team to underestimate the effort and complexity this project holds"

@HopelessN00b I don't even care about the $$. I just want it done right.

@ewwhite No. There's a reason infrastructure architects command a higher salary or rate. The experience is hard to come by.

make sure they have a test box of some form to test/fuck-up on

@ewwhite That's cool too. Say it and bill it anyway, and mail the extra $$ to me. :)

@Chopper3 Yeah, I explained that I often use other resources to either help with best practices or double-check my work.

for every production platform we have a 1:1 'Reference' platform for final testing and to help out more junior staff get practice in a lower-risk environment

I'm heading to EMC training given by a coworker later today... so even seasoned professionals need to learn.

it's MORE important for us

as we get listened to more, so need to be more responsible in our approached than juniors

EMC training? perks up Anything particularly nifty, beyond the niftyness of EMC in general?

This is an 85-user site... I told them a 2-3 server VMWare cluster on G7 hardware...

and they aren't sure they are ready for virtualization

yep, support you on that, do they need HA/DRS

@HopelessN00b VNX 5100/5300

it's the norm now, not the exception, has been for a few year

Oh, cool, we just got two of those in...

@Chopper3 I'd present Essentials Plus and Standard.

ZFS-based SAN

no idea at that end sorry, one-SKU-man me sorry

@ewwhite Well, I'd reassure them that it's much better and cheaper than the alternative of buying more physical boxes with more complex setups (mashing multiple services and functions onto an OS to save on buying a new server).

I tried.

@Chopper3 Bristol Uni, for a small charge, will allow students to hook their servers into the Uni cable if they want to use it for personal projects or game servers.

"You might not feel ready, but it really is the least imposing, easiest approach to your problems... gotta go with the best option, even if you don't always fall in love with the best available option."

@tombull89 nice

It's like they don't trust certain things... I suggested an Amazon instance to spin up a web server quickly...

@HopelessN00b need to do some more digging but I think my cupboard is safe :)

and they said, ""we'd rather have control of it... so we'll put the web server on a PC here until we make a decision on server hardware"

@ewwhite Well, ultimately, if they just keep saying "no" and ignoring your advice/recommendations, what else can you do? You can't force them to do it right... "you can lead a horse to water, but you can't make it drink" and all that.

@Iain Yeah, being on another continent and all... but if you eat out, especially at a chain restaurant started in America, you might wanna take a closer look at any packaging. There's probably something of ours in what you're eating.

@HopelessN00b You are right... it's just sad. And it makes me think that these are the types who are being left behind.

with outsourcing, cloud, etc.

@HopelessN00b OOh, what's the product

@HopelessN00b When I'm in NY next year I'll keep a look out

@Dan Food. Lot and lots of food of many varieties. Some of it's sold to consumers by retailers, some of it's sold to (for example) McDonalds for use in most of the 7.6 kagillion burgers they sell every day, and some of it's even sold to companies that rebrand it and sell it at a markup.

And maybe other ways that aren't immediately coming to mind.

Ah, right, thought it was packaging or somethi

How do I spot it?

@HopelessN00b @Chopper3 um, I'm trying to move my entire page file to RAM, can you help me figure it out?

@HopelessN00b Ah right, fair one

Big corp, then?

@BrentPabst Yeah. Hit yourself on the head with your RAM disk until hear a loud crack and end up with grey gooey-stuff on the RAM disk. Repeat as needed until death.

And depressingly, I'm actually editing more into my answer on account of his dumbass comment-replies. sigh I never learn.

@HopelessN00b He sounds like he needs to go work for a university so he count preach the new ways of the world

@Dan Yup, "mega-corp." Like I said earler, ~2 billion (USD) market cap.