|
MF_James posted:I can't tell you how many times the garbage toners will happily light up but data still will not pass through a jack, as soon as I have the tech punch the jack down again magically it starts working. Yeah if I'm paying someone to run cable, then they're gonna provide me with a cut sheet and certify each drop. If even a single one of those turns out to be hosed, they're coming back and fixing it. Net 30 means I have plenty of time to test them myself or at least find out accidentally if anything is hosed.
|
# ? Oct 26, 2018 23:12 |
|
|
# ? Jun 3, 2024 22:57 |
GnarlyCharlie4u posted:I mean honestly, I've never seen so many "oh my god I'm afraid of cables" posts in one place. Crimpin' ain't easy.
|
|
# ? Oct 26, 2018 23:23 |
|
I just want to know how the logistics works of wiring a couple hundred drops and getting them to fall into the patch panels in a way that means the numbers are in the same order as the drops are out on the floor.
|
# ? Oct 27, 2018 00:27 |
|
All this talk about cabling makes me very happy that Oregon requires a license for that sort of thing. Sorry, boss - I can't wire that office expansion, we'll have to get a licensed vendor to do it! gently caress punching down cables forever.
|
# ? Oct 27, 2018 00:43 |
|
Exodor posted:All this talk about cabling makes me very happy that Oregon requires a license for that sort of thing. So we're sending you for mandatory training...
|
# ? Oct 27, 2018 00:45 |
|
Neddy Seagoon posted:So we're sending you for mandatory training... I got asked to do something recently and I demurred because no certificate. To two hours later I get a receipt for a certification webex course registration emailed to me. Due date: next Wednesday. Course length: 8 hours. 'just finish it in your spare time lol' This should probably go in the pissing you off thread
|
# ? Oct 27, 2018 01:34 |
That would be my entire workday Monday if they pulled that poo poo.
|
|
# ? Oct 27, 2018 02:23 |
|
Thanks Ants posted:I just want to know how the logistics works of wiring a couple hundred drops and getting them to fall into the patch panels in a way that means the numbers are in the same order as the drops are out on the floor. If you're thinking ahead you're marking the cables as you're pulling them. It's not like you don't know where you're pulling them to. If you haven't done that, hopefully you left enough slack at the patch panel end that you can send out a PFY with a toner and work it backwards. If you haven't done that, hopefully no one actually cares about the order and you can just patch them in where they fit, then number the jacks on the floor based on where they ended up. Again a PFY with a toner helps, or if you have nice enough switches a laptop and something to view CDP/LLDP with. Even if you do mark as they're run though, $deity help you if someone in power actually cares about the order of the numbers on the floor. Inevitably somewhere is going to need a new drop and you either have to have drop 73 physically in between 14 and 15 or someone's getting a pile of bitch work.
|
# ? Oct 27, 2018 02:41 |
|
MF_James posted:I can't tell you how many times the garbage toners will happily light up but data still will not pass through a jack, as soon as I have the tech punch the jack down again magically it starts working. That's because toners aren't for that. Toners are for identifying and locating wires. Some of them, like Fluke's Intellitone 200, will tell you if a wire is open or something, but to actually prove that a cable performs to specification, you need a cable certifier. They test much more than wire map and will test for things like NEXT and return loss. They are fantastic and anyone pulling or testing data cabling should have one, but they are expensive, much more so than a toner. A really good toner is a couple hundred bucks. The cheapest good copper cabling certifier I know of is still thousands and thousands of dollars. I can terminate a cable on request that will pass a wire map test but won't meet Cat6 performance standards. The most important thing about a certifier is that a calibrated tester operated by a certified operator is unimpeachable. If you show a contractor a certified test result that says their result fails, that's the end of the story. There's no arguing, the cable failed. A reputable cable vendor will have one of their own and provide certification results at the end of the job. Don't hire someone to do data work if they can't or won't provide that certification. This is one reason I don't like to hire electricians to do data work; they usually don't have one, don't know what it is, don't have anyone certified to operate it -- not that it's difficult, especially for copper -- and definitely won't pay to get either one. Thanks Ants posted:I just want to know how the logistics works of wiring a couple hundred drops and getting them to fall into the patch panels in a way that means the numbers are in the same order as the drops are out on the floor. There's no magic to it, it's just careful planning. On a new construction project a data contractor will have prints to work from showing how many drops go where in each room, and a good one and will plan out what's being pulled where long before they actually pull any cable, and they'll have a port map for each closet. They will mark cables physically before pulling a bundle so they know which cable is which when they get it there, and there will be plenty of slack when they go to cut down the cables to length. After that it's just a matter of being meticulous about punching down the cables as planned. The order the drops are punched down in matters -- sort of. If I'm having issues with drops in a room, I would prefer that all the drops in a room be next to each other so that I don't have to hunt all over the panel for them. And ideally I would like adjacent rooms to be adjacent on the patch panel for the same reason. But after the initial installation, any new work is almost certainly going to go on the end of the next free panel no matter where it is. You can only be anal retentive about this stuff for so long before reality will intrude. (Make them follow the existing labeling scheme. Don't let them add BREAKROOM-1 and BREAKROOM-2.) Bonus reason I don't like to have electricians doing data work: their labeling sucks. Sometimes they just won't label at all, unless you put your foot down. Other times they'll hand-label it, which will fade over time and eventually you'll have no idea what's what. Labeling should be done with a machine labeler so they won't fade. (Once I had someone label in a way that the labels on each end weren't the same, which defeats the purpose of labeling. Yes, they were given a labeling scheme to use. They ignored it.) When you don't have a label, a toner will do the trick, but it's much nicer to know that you're looking for D1-A23 or whatever on the panel. guppy fucked around with this message at 11:10 on Oct 27, 2018 |
# ? Oct 27, 2018 11:00 |
|
guppy posted:That's because toners aren't for that. Toners are for identifying and locating wires. Some of them, like Fluke's Intellitone 200, will tell you if a wire is open or something, but to actually prove that a cable performs to specification, you need a cable certifier. They test much more than wire map and will test for things like NEXT and return loss. They are fantastic and anyone pulling or testing data cabling should have one, but they are expensive, much more so than a toner. A really good toner is a couple hundred bucks. The cheapest good copper cabling certifier I know of is still thousands and thousands of dollars. Sigh. What a timely discussion, annoyingly enough. I just had a project come up for my side gig, one that I tried to avoid, and gave them a fuckyou price to get them to go away, but they decided they still wanted it. The project is to reterminate about 60 drops (wires, not locations), into new patch panels, put in a four post rack and mount the patch panels / gear in it, and patch/wire everything. They're going to end up paying me about $200/hr (I hope, assuming I don't take forever - I'm normally a sysadmin not a cable puncher, and in case that comes out wrong, what I mean is I have utmost respect for those guys that can just line up the cables and punch them down bam bam bam - I'm not that guy because I do it so infrequently and the last major wiring project I did was the 48 drops in my house. In 2011). So I'm pondering using some of the money from the project to get a certifier, and yes, I know the entry level ones are $2-3k. My question is, after staring at reviews for a while, what would you recommend? The Fluke range appears to have been sold off to another company, and the few reviews of the newer models don't seem promising (poo poo like the thing can only store reports for 50 cables and it takes hours to download them to a PC). There's the Ideal Navitek, but reviews of that aren't great either. Any certifiers that are actually worth the money in that range?
|
# ? Oct 27, 2018 23:24 |
|
Have you looked at the Viavi range? The company is what became of JDSU and telcos use their DSL stuff pretty much without exception.
|
# ? Oct 27, 2018 23:29 |
|
Thanks Ants posted:Have you looked at the Viavi range? The company is what became of JDSU and telcos use their DSL stuff pretty much without exception. On the one hand: Thants! But on the other, their Certifier10G, which looks like the lowest-end copper certifier, seems to be streeting around $8k. Which is, uh, a fair chunk more than I want to pay. I'm basically wondering if there's anything just up from the Intellitone that will certify gigabit (don't even need 10g) and is less than $3k, that's worth it. So thants but no thants. (But no seriously, thants for at least alerting me to them, I hadn't heard of them before and it's good to know about at least).
|
# ? Oct 28, 2018 00:21 |
|
Still quite pricey but here's Ideal's certifiers. I have their "verifier" (i.e. pair and distance-to-break tester) and it seems quite good. Haven't tried the certifier on account of it being 40x more expensive.
|
# ? Oct 28, 2018 00:33 |
|
I have an Ideal Navitek NT Pro at work and it was very good for the price. It’s a tester not a certifier but the brand seems decent. The firmware started off reading the IP address that it had picked up backwards which was a bit of a bad first impression, but a software update fixed it. Thanks Ants fucked around with this message at 01:07 on Oct 28, 2018 |
# ? Oct 28, 2018 00:58 |
|
Hrm, maybe I just go with a tester rather than certifier, in that case. The LanTEK is a bit more in the price range, but since I don't plan to do this stuff often it probably isn't worth it. I just wanted something a little better than "yup it's electrically connected". The NaviTEK was already something I'd been looking at and the cost is definitely doable. It seems a bit crazy how much more the certifiers are. Don't get me wrong, I'd be the first to yell at someone for going "why should I pay $5000 for an enterprise switch when this $500 Zyxel switch has all the same specs", and I understand that good gear, especially when it's a specialized field, costs money. But on the other hand, there's lots of examples of specialized gear not really being all that special in terms of the actual equipment, and you're just paying a premium for it being specialized and because "that's just what it costs". And these days, testing that a cable can reach gigabit speeds seems doable with two laptops and iperf. What exactly justifies thousands more for what looks like a jumped-up Android tablet in a rubberized case with a network testing app?
|
# ? Oct 28, 2018 01:31 |
|
Well a lot of the good ones will pick up things like attenuation, cross-talk, wire breaks with distance to the break, cable length, etc... Way back when I did cable stuff, I remember having access to a Fluke validator and that poo poo was like magic all the data points it could give you.
|
# ? Oct 28, 2018 01:41 |
|
Super Soaker Party! posted:Sigh. What a timely discussion, annoyingly enough. Super Soaker Party! posted:Hrm, maybe I just go with a tester rather than certifier, in that case. The LanTEK is a bit more in the price range, but since I don't plan to do this stuff often it probably isn't worth it. I just wanted something a little better than "yup it's electrically connected". The NaviTEK was already something I'd been looking at and the cost is definitely doable. So the real answer to this depends on the terms of your contract. 1. If your contract does not specify that you are to certify the results then, well, certifying the lines at all isn't a necessary expense. But you'll never really know if your work is fully spec compliant without certifying. 2. If you just want to test the lines, you can look at a LinkSprinter (a few hundo) or a LinkRunner (~$1200). These were formerly Fluke branded, then sold off to Netscout, then just recently sold off to StoneCalibre. They will test for wire map and will literally test whether the line functions -- it will upload a report and email you a link to that report. If you got the report, then it uploaded. As far as I know it's only this line (LinkSprinter, LinkRunner, AirCheck, etc.) that was sold off, Fluke is still doing their Versiv units with the assorted copper and fiber modules. 3. Fluke is the gold standard of cable certifiers but they are no poo poo like $60K to buy a full kit and that's only copper. They are awesome but if you're gonna do this once or twice ever it's not worth that money. 4. Ideal is the only other company I know of making good cable certifiers, and that's the LANTek series. They are not as fancy as the Flukes but will give you a perfectly good certification result for $8k instead of $60k. I would personally accept either from my vendors. That's still a lot of money in an absolute sense but it's an order of magnitude cheaper than Fluke. 5. Here's what you might actually want to do: my understanding is that you can actually rent cable certifiers from a local vendor. For Fluke at least your results are not technically "valid" without an operator certification, but testing copper is not rocket science and you'll be able to test that your work performs to spec. It is very important, obviously, that you configure the thing for the right kind of test, cable type, etc. The other thing to check is that for Fluke at least the unit must be re-calibrated annually. For you to be able to trust the results, the copper module must have been calibrated within the past year, so if you're renting a unit, make sure it's within the calibration period. There should be a certificate indicating its calibration date. I don't know if Ideal does something similar or how necessary it is. Hopefully that's some help to you. As far as why cable certifiers cost so much, they can produce a really incredible amount of information that's very useful for troubleshooting, not just whether the line fails or passes. It will give you detailed reports for each test, and if the line does not pass, it'll tell you why. If there's a break in the cable, it'll tell you where the break is and on which wires. If you have too much crosstalk, it'll tell you which end is the issue. It will give you the margins by which you passed or failed each test. Most of all, it protects you and your work from problems down the road. Oh, you're having problems with that line? Well, here's the certification result from when I installed it, I can't do much about it if someone monkeying around in the data closet messed it up later. That said, there's more to cable installation than that. For example, the lines need to be supported properly on the back of the panel so that the weight of the cable doesn't pull wires off the panel over time. (Other stuff, like firestopping, you won't have to worry about if you're just doing the termination work.) There's a reason I hire pros to do cabling runs instead of doing them myself.
|
# ? Oct 28, 2018 02:33 |
|
Super Soaker Party! posted:So I'm pondering using some of the money from the project to get a certifier, and yes, I know the entry level ones are $2-3k. My question is, after staring at reviews for a while, what would you recommend? The Fluke range appears to have been sold off to another company, and the few reviews of the newer models don't seem promising (poo poo like the thing can only store reports for 50 cables and it takes hours to download them to a PC). There's the Ideal Navitek, but reviews of that aren't great either. Any certifiers that are actually worth the money in that range? I'm going to go ahead and say that your money is best spent, for this project, in just buying the time of someone who already has the certification gear and getting them to double check your drops. It may be a chunk of cash for a subcontractor's hourly, but its probably less than buying new,but on the flip side you might get lucky and could rent one/borrow one instead. (by which i mean think of it as 'rental with a chaperone')
|
# ? Oct 28, 2018 06:11 |
|
guppy posted:a bunch of helpful stuff Thanks! This is incredibly useful info. I definitely was wondering about renting one of these things, because yeah, I don't really want to spend $8k much less $60k for what's essentially a one-off (I don't plan to do much more of this kind of work if I can help it). I'll take a look at what that would cost around here to rent, but frankly it may just end up that I get a tester like the NaviTEK, which I WOULD in fact use other times than just this project, and call it good - the install I'm doing is not for a particularly demanding client so I don't think they'll be expecting a certification or anything more than basic test results. I suppose all that information could justify the specialized chipset / test equipment and cost inside the certifier - still does seem outrageous especially for the Fluke level, but I suppose if you're doing drops day in day out it pays for itself pretty quickly (especially with the ability to say later "no gently caress off it certified fine whatever problem you're having now is not my responsibility"). TheParadigm posted:I'm going to go ahead and say that your money is best spent, for this project, in just buying the time of someone who already has the certification gear and getting them to double check your drops. Yeah I'm definitely coming to agree on that point, but I'll talk to the client and frankly maybe even say I'll do basic testing but they need to pay for someone to certify after the install, with any fixes charged at extra hours at that point. I made it clear up front this was not really my usual work and that I'm not a wiring contractor, but I guess even my "go away" price was still cheaper than whatever other quotes he got, which is...ummmm. Guess there's something to be said for being able to charge $boat in the Bay Area. Something to keep in mind for future.
|
# ? Oct 28, 2018 07:58 |
|
Super Soaker Party! posted:I made it clear up front this was not really my usual work and that I'm not a wiring contractor, but I guess even my "go away" price was still cheaper than whatever other quotes he got, which is...ummmm. Guess there's something to be said for being able to charge $boat in the Bay Area. Something to keep in mind for future. Ask what the other quotes were while you're at it - even if its after the fact, or whatever.
|
# ? Oct 28, 2018 20:12 |
|
GnarlyCharlie4u posted:....Also, having to use those terrible through-end rj-45 connectors. Those things are the worst. Even if you have the special tool designed for YOUR BRAND of end, it still won't do a perfect job every time and you'll wind up with lovely crimps.... Not sure what problems you are having with them... But having crimped more than 600 of the things, I can count on one hand the amount of non-working crimps I have had. My only problem with them generally comes down to the blade.. After about 100 crimps or so, the blade is quite dull and needs replacing.
|
# ? Oct 29, 2018 14:16 |
|
I get bad crimps, I also start to lose feeling in my fingers lining up/straightening the wire ends over and over again because this time, hopefully, one of them won't jump the channel or not go all the way to the end.
|
# ? Oct 29, 2018 15:28 |
|
I just got off a 20 minute support call where the user sat there and fidgeted with the mouse the whole time i was remoted in and trying to do poo poo. I’m kind of bothered by how much it infuriated me.
|
# ? Oct 31, 2018 18:49 |
|
I'd be tempted to say something like "huh your mouse is still moving, must be a glitch - could you unplug it for me?"
|
# ? Oct 31, 2018 19:01 |
|
Security team: Agrikk you need to do a thing. This is critical that you do a thing. Agrikk: Uh. Okay. I'm happy do do a thing. Can you give me directions on how to do A Thing? Security Team: [To Boss' boss and Agrikk's boss]: Agrikk hasn't done a thing yet! Agrikk's boss' boss: What the hell? What is this? Agrikk's boss: I have no idea. Agrikk? Agrikk: All I know is that I got pinged to do a thing and I don't know how to do a thing. I asked for their help and they escalated to you guys. Agrikk's boss' boss to security team: What the hell? Please explain to Agrikk how to do a thing so every one in this chain can stop getting pages every 60 seconds. Capable guy on Security team: Agrikk, please do X, Y, Z to resolve this. This all happened because of <reasons> and everyone is getting jammed up with this Thing. Agrikk: Okay I've done a thing and have received approval code 1234567 that it has been approved. Capable guy on security team: Thank you. Security Team: Agrikk, please give us an ETA on when you will have this work done. Agrikk: I have completed the work. If you looked in this very thread in this case you will see I have received approval code 1234567. Security Team: Agrikk it has been fourteen days since this ticket has been opened. Please provide an ETA on when you will have this work completed. Agrikk: I have completed this work with approval code 1234567. Security Team: Agrikk it has been 21 days since this ticket has been opened. Please provide an ETA on when you will have this work completed or we will be escalating this to your manager. Agrikk: Escalate away. If you cannot read a ticket I cannot help you. Security Team: Agrikk, it has been 28 days since this ticket has been opened. We have escalated this to your manager and your skip-boss. Agrikk: There will be no ETA. There will never be an ETA since this work has already been completed. Agrikk's boss: Agrikk, have you done this work? Why am I getting paged into this thread? Agrikk: Read the thread. Agrikk's boss: Oh. I see. Carry on then. Agrikk's boss' boss: Agrikk and Agrikk's boss, why am I getting pages about this Thing? Agrikk's boss and Agrikk together: Read the thread. Agrikk's boss' boss: Oh. I see. Carry on then. Security Team: Agrikk this issue has been open for over thirty days. Can we have an ETA on this work? Agrikk, Agrikk's boss and Agrikk's boss' boss: No Agrikk fucked around with this message at 19:21 on Oct 31, 2018 |
# ? Oct 31, 2018 19:18 |
|
Security team needs to be re-orged to only "securing" that that vending machines are stocked.
|
# ? Oct 31, 2018 19:44 |
|
why they gotta do a thing
|
# ? Oct 31, 2018 19:51 |
|
So it would appear that InfoSec teams are the same across the board. Got it. The main issue I have with ours is that they vet every firewall change request before it goes through the networking team, and they have absolute veto power. Ok, that's not the issue. The issue is that the person involved in vetting the requests only knows how to follow a checklist and will hold them up if something doesn't line up with the checklist. A request for clarification will then be sent. Ok, that's not the issue. The issue is that when the person asks the question about the item that doesn't fit the checklist exactly. they will send the request for more information after hours, so the firewall change will be delayed at least an additional day. Ok, that's not the issue. The issue is that when I calmly asked them to request information during normal business hours so requests don't get delayed (typical turnaround without questions is the next morning), they responded only with "our SLA is 3 days".
|
# ? Oct 31, 2018 20:42 |
|
We inherited some bad technical decisions and people seem to be leaning into them, and nobody actually involved day to day thinks it's a good idea, but none of us have the will to fight it. We're hoping that security will do us a solid and fail the approach based on the lovely codebase we're having to work with. Hopefully they won't just rubber stamp it. Security concerns are the only things that get any traction these days.
|
# ? Oct 31, 2018 20:49 |
|
Can't remember if I told my InfoSec story from a few years ago. I'd just gotten a really nasty dressing-down from my boss, complete with yelling and threats, so I figured I was on thin ice. Came to work the next day, and my computer was gone from my desk. A coworker said, "Oh yeah, some IT guy showed up and just walked off with your laptop. Didn't say why." I started packing up my stuff: usually the computer seizure comes right before the escort off the premises, right? And I can't even check my email, so I expect this is it. Since I have little to lose, I call up one of my buddies in IT to say goodbye and ask him to find the ticket to see how how bad it is. "Huh--turns out InfoSec suspects there's malware on your machine and they wanted to examine it. You don't appear to be fired." Turned out the Chrome mouse gestures extension I had installed was periodically sending usage statistics back to its home server, and that set off some alarm on our web proxy. Which is fine and all, but you'd think a quick phone call would be in order. "Hey, we're deeply concerned about a potential threat on your machine. Here's a loaner while we examine things, and let's just confirm you don't have any critical deadlines today. We'll have your machine back by 3!" Nope, let's just send a tech to confiscate your machine.
|
# ? Oct 31, 2018 21:06 |
|
Johnny Aztec posted:Security team needs to be re-orged to only "securing" that that vending machines are stocked. Pretty sure they have to have a BSC and an SSC before they can do that.
|
# ? Oct 31, 2018 21:24 |
|
Johnny Aztec posted:Security team needs to be re-orged to only "securing" that that vending machines are stocked. Kind of a fox-guarding-the-hens situation, isn't it? Let them guard the office plants.
|
# ? Nov 1, 2018 03:01 |
|
The Macaroni posted:Can't remember if I told my InfoSec story from a few years ago. I'd just gotten a really nasty dressing-down from my boss, complete with yelling and threats, so I figured I was on thin ice. Came to work the next day, and my computer was gone from my desk. A coworker said, "Oh yeah, some IT guy showed up and just walked off with your laptop. Didn't say why." Yikes, the worst we ever got was a panicky E-mail because one of the seniors ended up finding a canary of theirs trying to figure what the gently caress is this unknown thing on our network. The cherry on top though is the conversation that followed; InfoSec: "So we've seen one of you has tried to access one of our secure hidden devices blah blah blah" Senior1: "Oh yeah that's just X hacking" InfoSec: "!?" Senior1: "Wait wait sorry the guy's surname is Hacking, he discovered this unknown device yesterday and tried to figure out what it was" Senior2:
|
# ? Nov 1, 2018 13:10 |
|
Johnny Aztec posted:Security team needs to be re-orged to only "securing" that that vending machines are stocked.
|
# ? Nov 1, 2018 14:24 |
|
Collateral Damage posted:Security teams are what you end up with when you give a group of people a singular responsibility without requiring them to weigh the business needs into it. Like a human paperclip maximizer. I know this is more of a joke, but I'd say this is more true of Security Analyst teams, or security C levels. Security engineers can be super effective in decreasing overall workload by providing good, non-invasive security solutions that prevent large threats from causing downtime. If your security team doesn't actively own and manage products or tools in the environment, they're doing it wrong.
|
# ? Nov 1, 2018 15:50 |
|
Exodor posted:All this talk about cabling makes me very happy that Oregon requires a license for that sort of thing. I'm making both of my guys get certified just because of that :p
|
# ? Nov 2, 2018 19:19 |
|
Agrikk posted:
" Please see attached documentation: https://www.amazon.com/Hooked-Phonics-Beginning-Reading-Ages/dp/B0012H8U9S/ "
|
# ? Nov 3, 2018 00:37 |
|
Che Delilas posted:" Please see attached documentation: https://www.amazon.com/Hooked-Phonics-Beginning-Reading-Ages/dp/B0012H8U9S/ " There is absolutely no doubt that the next time I get an update request on this ticket I will be linking this book. This is absolutely solid.
|
# ? Nov 3, 2018 22:19 |
|
Yay, I get to post in this thread now. A series of tickets came in: quote:Hello Hollow Talk, quote:Hello Hollow Talk, quote:Hello, I actually get to reply here, in a markedly more polite form that basically still said: "Hi customer, they are probably missing because your test files are broken and because I can't just 'disable' the core integrity checks of this whole ETL-process." quote:Hello,
|
# ? Nov 5, 2018 21:33 |
|
|
# ? Jun 3, 2024 22:57 |
|
Welcome to "Life With Devs"
|
# ? Nov 5, 2018 23:27 |