« first day (698 days earlier)      last day (4298 days later) » 

1:00 PM
if you remotely think someone would / could screw you, then you probably shouldn't be working with them
 
If they're relying on you for part of their income, you can't just take some guy's word.
Um...and if you died?
 
i dont think that of my partner
 
morning
 
That's your business partner.
I'm talking about clients working with you for a product.
If you're the only one doing everything and you're incapacitated...stroke, car accident...
I'm screwed for having bought your product.
 
good point
 
1:01 PM
It just doesn't scale.
And when you're a school or medium business relying on your stuff...that screws them majorly.
So they look at things like the probability you'll be around in five or ten years.
 
thats not the market, but i see the point
 
even on go live day, to me its still just an idea. once i have students using the website on a regular basis is when this is no longer a garage startup and we have something real. after that, for sure, sometime will need to happen to secure its future
 
@BartSilverstrim: the bus rule? ;p
 
More or less.
Although in NYC I think of it as the subway rule.
Keep the PDA up on the platform and someone might accidentally give you closeup view of the rails.
 
1:04 PM
(ugh, sorry about the dupes, guys, Its a glitch due to the epically craptastic internet dongle i'm stuck with here)
 
@MDMarra Another nice Mountain Lion discovery... you can have multiple backup drives in Time Machine.
 
I heard ML also encrypts backups to time capsules.
 
I believe it can.
 
that and the common carelessness of using same username / password / cc# / etc... across multiple sites
 
I think I need to do an edit pass on that canonical raid question. It's advice is fine for servers, but not relevant for storage
 
1:10 PM
@Basil Right, it applies to direct-attached or internal storage. Most of those rules are different when dealing with enterprise storage.
 
@ewwhite essentially, yeah.
 
I'd add a disclaimer.
@lsiunsuex So what should people do?
 
There was an announcement yesterday that Amazon changed their practices to stop this.
An announcement that someone discovered they changed their policies anyway. Amazon wasn't advertising the change.
 
At 5:02 p.m., they reset my Twitter password. At 5:00 they used iCloud’s “Find My” tool to remotely wipe my iPhone. At 5:01 they remotely wiped my iPad. At 5:05 they remotely wiped my MacBook.
 
They should just use two level authentication, a simple text msg with a pin code would be suffice
 
1:14 PM
the multiple cc thing is a bit extreme. CC's have gotten me in trouble in the past and i try to only keep 2 at max. 2 factor auth is a good idea like @ColdT said but that needs to start with the service provider. i also don't think service provides (even apple or amazon) should keep ANY cc info on hand, last 4 or otherwise
 
Other things can mitigate the risk too. Like encrypting your notebook and not relying on third party wiping ability.
 
And offline backups... I just went through a lot of pain to keep my Time Machine backups that reach back to November 2011 intact.
 
Yes, he readily admitted he was an idiot there.
 
which for this dude, could have been a $100 USB2 hard disk.
 
Morning all
 
1:17 PM
Morning.
 
the hack was simply a grab for my three-character Twitter handle. That’s all they wanted. They just wanted to take it, and fuck shit up, and watch it burn. It wasn’t personal
 
last time i checked, the ATM receipt had my name on it and the last 4 of my card # - this could have easily taken place locally and not remotely also
who doesn't have an itunes acct? the info could have been gleamed from a garbage can outside of a bank, let alone across the internet
 
lol. mine's on a credit card number that expired 5 years ago ;)
 
just so happens this was an "internet figure" and he had a lot to loose.
 
<- doesn't have an itunes account. :P
 
1:24 PM
@lsiunsuex that's actually a good point, never thought of that before with the ATM receipt
<- doesn't have any Apple products altogether :P
 
Hey, could I bother you guys for a quick sanity check? Or since I'm going to do it anyway, please excuse the quick sanity check. What do you guys think of the below statement?

"100TB is not a pain to back up. It is doable in ONE rack server (see: backblaze storagepod) and all you need is a decent LOWER END tape storage library to back it up. I know a lot of smaller shots that run sizes like that or larger. Not something you do with "home level equipment", but not something terribly hard to do either."
 
But yeah, those last 4 digits are in plain text all over the place. Most order confirmations and invoices will display that much as a reminder of which card you used.
 
don't think lower end tape storage would have the capacity to use LTO5 tapes or higher?
 
@SmallClanger: I bought an ipod years ago, the account is primarily for the crappy album art feature.
 
hand full of LTO3 would do the trick but need a magazine for it which unlikely found in lower end tape backup
 
1:28 PM
@ColdT Not to mention that even with LTO5, you're talking 67 tapes uncompressed. (Or maybe 40 with compression. For one backup. Doesn't sound like a low-end solution to me.)
 
@HopelessN00b Backblaze are not a good example. They have their own architecture that lets them use such cheap boxes.
 
this isn't a problem with apple / amazon. this is a problem across the board with our culture. people are in general "trusting" of each other. goto a restaurant tonight and pay with your card. the waiter who probably looks like a dumb fuck and you feel bad for cause he has a perceived shitty job may be one of those hackers. now he has your name and last 4 of your CC# and pulls the same stunt this asshole did.
 
oh crap 100TB, guess that's not going to work
 
(or i'm a crazy, over worked paranoid f' )
 
@SmallClanger Yeah, and I'm not sure that even they'd pull it off all that cheaply, or easily.
 
1:30 PM
@HopelessN00b tape backs just don't seem feasible for 100TB backup..
 
Right, right. I mean, it can be done, and I've done it, but it required multiple tape libraries and large pain in everyone's ass. And absolutely no "lower end" gear.
 
100TB is not a trivial thing to backup
 
hell yeah pain in the ass without a doubt, and agreed, no way lower end gear.... anyone who want's to backup 100TB need some serious cashflow, else their business isn't going anywhere
 
what are the chances that that 100TB are live data though?
that could be 10TB of real data, then a bunch of copies
or 100TB of live data, most of which is archive or stagnant media
 
@Basil if your talking about simulation data, that can exceed easily past 10TB
 
1:34 PM
ANyone looking at that volume of data should KNOW these things
Otherwise, they're unqualified to be managing it.
 
@Basil, it was actually a response to one of my answers (by a real high rep user here) to some guy asking about distributed RAID5 system to store 100TB of data.
 
@HopelessN00b Link?
 
I've never had to deal with that amount of data, but if I did, I'd know where to go to find the solutions.
 
@ColdT Yeah, I wasn't sure about the source of the discussion
 
Wanted to make really sure I wasn't wrong before challenging it, considering the source.
 
1:36 PM
@Basil i doubt that OP is talking about simulation data either, was just a thought. But OP must be absolutely mad thinking he wanting it on a distributed RAID 5
 
4
A: Best way to safe store data on several servers?

HopelessN00bIt sounds like your requirements are wildly unrealistic. Firstly, 100 TB of data is a lot. Do you really need all of it to be available at the same time? If so, you need to be looking at a lot more than just spreading it out over several servers. There are networking and access considerations...

 
Got to completely agree with @ewwhite here, anyone with that volume need experience and certainly not for the lightweight
 
And even with research, I'd need a pro's advice... because I'm sure strange things happen at that scale of data.
 
you'd need some pretty experienced storage specialist dealing with 100TB!
 
@TomTom: Are you sure you didn't misread the question or answer? 100TB. Not 100GB. — Basil 6 secs ago
 
1:39 PM
@basil - you know how to do that?
 
Got a quick query about PsExec & xCmd, but not quite sure it's worth a post. In short, are they the same thing? Does one use the other? Or are they completely different and independent tools?
 
@Iszi It's worth a post :)
If someone answers it here, google won't catch it
 
@Basil i think that was a misunderstanding... unless @TomTom has dealt with that size of data before
 
@ewwhite link a comment? Yeah, just right click the "just now" link
 
@Iszi definitely worth a post. I'll answer it
 
1:41 PM
@Basil Contrarily, in my experience, it will. I'll go ahead and post though.
@Basil @JustinDearing Here, or Super User?
 
@Basil No, do you know how to handle 100TB easily?
 
@Iszi here definitely.
 
@Basil Oh that would make sense. If he was thinking kind of in line with reality, and thought the discussion was about 100 GB, that would totally explain everything. I feel a lot better now, thanks.
 
@ewwhite Oh, not easily: all my data is structured, and 100TB of structured data is a hell of a lot harder to manage than 100TB of audio
but yeah, I am on a team that manages over 1000TB
 
@Iszi in general, PSEXEC supports only one remote command on the remote machine, xcmd allows u to execute internal commands on the remote machine. If it was me, i would use xcmd
 
1:44 PM
@HopelessN00b He probably didn't misunderstand, and actually thinks that 100TB is tiny. He's a strange dude. Usually wrong, and almost always advocating running critical applications on mobile phones. In short, you should always ignore anything he writes, except to post it here for the lulz
 
0
Q: PsExec and xCmd

IsziI've recently come into a bit of confusion regarding PsExec and xCmd. It would be great if someone here could help clear this up, preferably with some form of external reference to given for validation. I understand that PsExec and xCmd are tools that can be used to execute commands remotely on...

^^ Ping @Basil @JustinDearing @ColdT
Thanks, guys.
 
@Iszi Sorry, I don't even know what those letters stand for :P I'm a storage guy.
 
@Basil Seriously? Guy with 24,000 rep? Wow.

ServerFault might be an even stranger place than I thought.
 
lol
people specialise
 
Sure, but... even specialists should be able to grasp the bare-minimum fundamentals in other areas... or know well enough not to throw out insane opinions on areas in which they're not specialized.
 
1:50 PM
@HopelessN00b it's either a complete misunderstanding or the guy really knows how to easily backup 100TB, one or the other, and if not, then damn this place is strange like Diagon Alley!
 
@ColdT It's not that he knows how to backup 100TB, for sure. I know how to backup 100TB. Not that I'd ever want to do it again, but... anyone who knows how to do it isn't going to be saying it's easy and achievable with lower end gear.
 
hmm that's true
 
@HopelessN00b You can get high levels of rep even if your average quality is low. He's the poster boy for that.
 
There's LOADS of stuff you guys talk about all the time that I'm clueless about - for instance I'd consider myself intermediate at best regarding puppet/chef, know nothing about linux-based DBs other than Oracle and probably have to use google twice a day at least to work out the best linux command to use for the more advanced stuff - I'm fucking ace at a few other things thought right ;)
 
@Chopper3 how would you suggest backing up 100TB worth of data?
 
1:59 PM
@Chopper3 And I don't know databases or Windows...
 
@Basil Gotcha. That helps a lot, though, I probably should have been able to figure that out already, considering my two highest voted answers are complete shit. (Well correct, but painfully obvious.)

@Chopper3 Yeah, but you don't post about shit you're clueless about as if you're an expert on it... do you?
 
@ColdT this would be a function of what changes and how often...
 
@Chopper3 So how extensive are the google results for less?
 
@ColdT piece of PISS - put a netapp on another site with 5-15 minute snaps and push it to that
@ScottPack haha
 
haha @Chopper3 i should add 'cheap solution' to the sentence!
 
2:01 PM
@ewwhite I know MSSQL and Oracle enough and I'm say intermediate at most Windows things, pretty good at a few bits (say terminal services, clustering, FC stuff)
@ColdT Buy a FAS 2240 then?
 
@Chopper3 You're a god at some stuff. Now build me a YouTube killer on this laptop...
2
 
@ColdT There is no cheap solution. If you're dealing in that much real data, you'd have to expect to spend on robust or reliable infrastructure.
 
@HopelessN00b hopefully not no
 
hey @Chopper3: I'm rewriting the raid 5 part of our canonical raid answer
 
oh yes, have fun
 
2:09 PM
Mind if I post it here so you can pick it apart before I submit?
 
sure but I'm on a flight in a bit
 
You own the plane, can't they wait?
 
I wouldn't mind taking a look, not that I'd be able to provide expert input on it. Unless the rewrite is "admins who use this RAID level in production systems go to the special hell."
 
richard branson owns it actually
 
Nice...
 
2:11 PM
#RAID 5
### Good when: You want a balance of redundancy and disk space or have a mostly random read workload.
### Bad when: You have a high random write workload or large drives.

RAID 5 has been the most commonly used RAID level for decades. It provides the system performance of all the drives in the array (except for small random writes, which incur a slight overhead). It uses a simple XOR operation to calculate parity. Upon single drive failure, the information can be reconstructed from the remaining drives using the XOR operation on the known data.
 
You own Richard Branson, can't he wait?
 
Time for a morning bike ride. Enough SF chat...
 
Is there a meta post about the raid question?
61
A: What are the different widely used RAID levels and when should I consider them?

Mark HendersonRAID: Why and When RAID stands for Redundant Array of Independent Disks (some are taught "Inexpensive" to indicate that they are "normal" disks; historically there were internally redundant disks which were very expensive; since those are no longer available the acronym has adapted). At the mos...

 
@Basil Why no mention of RAID-3? Did someone decide that's not ewven good for hosting video?
 
@JustinDearing Read the title ;)
widely used RAID levels.
Not "every fucking RAID level anyone's ever thought of that two people still use"
 
2:25 PM
I don't like a lot of things about that answer. Why is RAID-1E in there?
If they're including RAID-Z, they should also include RAID-DP
 
what's RAID DP?
 
they're equally suited to be in the answer (which is to say, neither belong there)
 
I've heard of RAIDZ, never RAID-DP
 
netapp's raid
raid Z isn't raid, it's ZFS
just like RAID DP isn't RAID, it's netapp
 
wait
raidz is raid
 
2:26 PM
what about x-raid lol
 
it's only available on servers, and I don't think it's on the standard
 
Gah, everyone's said 'raid' too much and it's gone all weird.
 
the difference between the two is that you need a netapp to use raid-dp. you can run raidz on commodity hardware
 
The standard RAID levels are a basic set of RAID configurations and employ striping, mirroring, or parity. The standard RAID levels can be modified for other benefits (see Nested RAID levels for modes like 1+0 or 0+1). Other, non-standard RAID levels and non-RAID drive architectures provide alternatives to RAID architectures. RAID levels and their associated data formats are standardised by [http://www.snia.org/ SNIA] in the [http://www.snia.org/tech_activities/standards/curr_standards/ddf/ Common RAID Disk Drive Format (DDF) standard]. RAID 0 A RAID 0 (also known as a stripe set o...
Although all RAID implementations differ from the specification to some extent, some companies have developed non-standard RAID implementations that differ substantially from the standard. Non-RAID drive architectures—configurations of multiple hard drives are not referred to by RAID acronyms. Double parity Now part of RAID 6, double parity, sometimes known as row diagonal parity, http://www.netapp.com/library/tr/3298.pdf 3298 like traditional RAID 6, features two sets of parity checks. Differently, the second set is not another set of points in the over-defined polynomial which charac...
@MDMarra Commodity hardware or not, it's not standard.
 
I didn't say standard anywhere
Just widely used
 
2:28 PM
Also, I think the raid levels should be in descending order or importance: raid 5, raid 10, raid 6, then the rest.
 
It's grouped in order of mirrored raid levels and parity based raid levels now
seems like it would be easier for a raid noob to follow that way
 
I know, but most people are just looking to see whether they should take raid 5, 6, or 10
 
2:45 PM
@Basil 60!
 
@Basil RAID5 should only be at the top if it comes with a big warning not to use it.
Otherwise, bottom, since it's almost completely non-important from a syadmin standpoint.
 
@Basil, While I agree the R5 section was a bit bloated, I'm concerned that the reworded warning about R5 rebuild times seems very gentile. Not that I contest it's accuracy, but plenty of idiots will be reading it.
 
^This.
 
@ChrisS It's factual, and to be honest, raid 5 has gotten a incomprehensibly bad rap here. I get it- if you're using trash drives, you need more redundancy, but that's not the case normally.
 
RAID-Z may be confined to ZFS currently, there's no reason that it must be (other volume managers could use it), and I know I've talking about the possibility of HBA accelerated RAIDZ (though no vendors are currently talking about it).
@Basil You've been talking with a few people who have misconceptions and poor experiences with RAID5. I'm certainly a proponent of it in the "right" situations.
 
2:56 PM
@ChrisS Assuming you have a cache, and assuming you're using 10 or 15k RPM drives, there's no performance or reliability reason to use raid 6 or 10
 
Oh, damn. Pro-tip, "Server 2003 **DISC 2**" CDs aren't bootable and cannot be used to install the OS.

Holy hell, I sure hope no one else noticed that.
 
@HopelessN00b raid 5 doesn't need a warning other than to use it with cache and fast drives. It only burns people who use cheap sata with it.
 
@BartSilverstrim Awesome! Now the police can be more effective and really crack down on people who produce those dangerous drugs that are corrupting America.
 
@Basil Everyone else already knows that, though. The people who don't know that, the audience for the question, are in desperate need of that warning. In fact, for hose people, that audience, it needs to be bolded, bigger, and in red font.

(Not that I'd trust it for anything production anyway.)
 
2:59 PM
In other news, the big city I live in a suburb of, is voting in November to decriminalize possession of Marijuana (up to a very reason limit)
 
@HopelessN00b You don't use cache and 15k drives for production?
 
@BartSilverstrim There are so many grow-ops around here that they've given up looking for them.
 
@Basil I do. Someone who needs to ask the question... well, do you thin it's safe to assume they're using the quality hardware they ought to be?
 
@HopelessN00b Of course, every shop is going to employ new people at some point. In any case, the answer is true, regardless of the sophistication of the reader.
 
(And to be fair, it's not always the sysadmin's fault. I've been in enough small shops to know that the same guy who spends $200,000 on a car is often the first guy to demand you save $30 on a hard drive.)
 
3:03 PM
@HopelessN00b WTF are you playing with a 9 year old installation disc for?
 
@Ward But they may find my nuclear accelerator project if they audit my records...
 
in that case, the answer clearly says "don't use raid 5 if you're not using 10 or 15k RPM drives"
 
@ChrisS Remote Repair install over ILO. System crash at a small remote site we haven't gotten around to on the hardware refresh yet.
 
@HopelessN00b Man, does that sound familiar.
 
I got a question about book/resource recommendation. The mods closed my question and said I should ask it here. I am looking for Books/Resources which explain TCP/IP implementation. I only know about:
1) Internetworking with TCP/IP: Vol.II, Design, Implementation, and Internals
2)TCP/IP Illustrated, Vol. 2: The Implementation
but these are quite dated. I am particularly interested in Open source implementations. Any ideas?
 
3:05 PM
@Bruce Depends. What level of TCP/IP knowledge do you have, and what are you trying to learn about, more specifically?
 
@Bruce "these are quite dated" <-- You're under the impression that IP has changed in the last 40 years?
 
@Chris: TCp has
 
Wow... I'll get it right one of these times.
 
3:08 PM
@Basil I dunno, double disk failures are trouble.
 
@Bruce Other than a few rather minor additions and security recommendations TCP/IP hasn't changed since 1981 tools.ietf.org/html/rfc793
 
@Basil I also think sysadmins who've been around the block a bunch where RADI5 was used (even on server-grade hardware) have seen a much higher array failure rate during rebuild than you'd expect. I know I sure have.

Hence the really bad rap for RAID5.
 
knock on wood I've never had a R5 array fail yet.
 
Maybe God just hates me, then.
 
@MikeyB They don't happen much when you're using 300GB 15k RPM drives
@HopelessN00b on servers, maybe. It's still the best choice for storage (unless it's something like 3par or XIV)
 
3:14 PM
rebuild is fkd up on r5 thats one thing i do know!
 
@Basil True. Not everyone can afford the sexy sexy small 15K drives though.
 
apologies for the language for under age!
 
@ColdT It's not that bad - the trouble comes when you're already hitting the array hard and rebuild takes days.
 
yes that would be DBs @MikeyB
 
@MikeyB Like I said, if you have 10 or 15k drives and a cache, raid 5 is the best raid for them
 
3:16 PM
@Basil That kinda depends on the server, doesn't it? A lot of ours use higher capacity 10k or 7200 RPM (near-line SAS). RAID5 those and you're begging for trouble.
 
@HopelessN00b high capacity drives don't belong in a raid 5.
@ColdT My rebuilds are 6 hours.
 
Indeed not. But a lot of server-grade hardware thees days is high capacity drives.
 
@HopelessN00b We use 10k Midline SAS drives for our main VM storage. 2GB BBWC and a bit of read-ahead tuning and it easily saturates the interconnect.
 
@HopelessN00b Not if you want performance.
 
I see a lot of 900GB 10K drives go through.
 
3:18 PM
had a major problem with our splunk server this morning, I didn't want to deal with it as I'm off on holibobs soon, so someone looked at it, turns out the company that does our media recommendations service changed their xml responses, up until now they were around 1k-2k each, as of yesterday when they did an upgrade they're now 24k-32k - as an xml response to a query - ffs!
 
Emotional rollercoaster: "MS have a KB article detailing the exact problem I'm seeing. Woo!" - "It has been fixed. Yay!" - "In an update rollup I installed and superceded over a year ago. FFFFUUU..."
 
@MikeyB We're planning on using raid 6 with those. Still not raid 10, though.
 
@Chopper3 That wreaks of auto-generated code and XML output. =]
 
It really burns my tits that we've had an XIV sitting here for the last couple years and not enough power to run it. Shoemaker's children syndrome, I tell you.
 
@MikeyB I don't like XIV, myself. It was interesting when they launched, but 3par is everything they are and more
 
3:20 PM
@ChrisS problem is we get a few hundred a second - and they all go to splunk, so the volumes went up (a lot)
 
@SmallClanger and now you get to report it as a regression! (that's when the coaster goes off the rails, over the side, down the ravine, through the nest of badgers, and into the lake - right?)
 
Angry badgers with herpes, too.
2
 
anyone have experience with ustream ?
 
@lsiunsuex I'm aware of its existence, and have watched things on it
 
3:33 PM
@lsiunsuex I do
I use it for a podcast
for now, at least
 
reliable? any way to avoid the ad's ? (paid acct maybe?) any comments on # of viewers?
i saw they had an api - maybe i can embed it in my site
 
It's terrible. Worst fucking website on the internet. They replace your stream with ads every 15 minutes, and their quality isn't anything to write home about.
I embed it on my site
 
shit
 
I'm switching to google hangouts asap, though
 
alternatives?
so google hangout
 
3:35 PM
justin.tv, google hangouts
do you need video too?
 
ultra music festival live streamed to youtube back in march
video and audio - hopefully 50 viewers from across the country - more the better (gonna email blast my list of students)
 
Does it have to be live?
You can do youtube live now, I think
 
preferably - if not possible, yeah, i'll record it and through it on my FMS server afterwards
 
otherwise, google hangouts
 
we bumped the go live to this monday cause we're hosting an event for our local college's dental students on the original live date (1 week from monday). although the event is local, the content is not - i can blast my email list that a live event is taking place and they could watch it. the exposure would be huge
 
3:39 PM
You should play around with google hangouts- it's fairly intuitive, and allows you to post straight to youtube for posterity
 
google hangout is limited to 10 people - so thats no good
 
10 presenters- unlimited viewers
 
looking at it. thank you.
 
welcome
 
Someone's list of the top 10 "WTFs in Video Games" youtube.com/watch?v=oO3WdL6BSaI&feature=related NSFW
Some of those are really funny...
 
3:43 PM
Uh.... gawddammit, this environment is going to drive me insane. Insaner.

I just saw a server running 2003, that wouldn't into Windows boot because of a missing/corrupt system path directory boot up normally with the CD mounted over ILO. Brain hurting... not comprehending... the hell?
 
@HopelessN00b Seems legit to me. Your OS on disk is corrupt, no reason you can't boot from CD (even virtual CD)...
 
Only it didn't It booted from the hard drive. The one with the corrupt OS image on it.
 
@HopelessN00b . . .
it's Windows? :-)
 
Someone needs to drop less acid before coming into chat . . .
 
I'm not on acid. :p

Would kind of explain a lot, though. Wonder if someone spiked the soda fountain...
 

« first day (698 days earlier)      last day (4298 days later) »