|
DigitalMocking posted:poo poo pissing me off: Bullshit replies to a trouble ticket. You see this a lot in telecom stuff. Either ping will be turned off or it will drop every 3rd packet as some sort of rudimentary DDOS protection.
|
#
?
Jan 8, 2016 15:17
|
|
|
|
| # ? Jun 8, 2024 14:25 |
|
|
Yeah, to be fair i've seen that too. I love when my boss just drops random people in my office - "oh hey these guys maintenance our racks in the building!" Okay. "You're going to work on zabbix with them now!" Noooo... i'm going to work on the lovely phone system that you threw a hissy fit about and called me at like 7 am for. The phone system that required no intervention from you with the plan i'm currently carrying out. "Oh uh... well i'm going to test some wiring" K please do  Its just that mix of poor planning, rushing at the last second to save face, and promising the world to everyone that has worn thin with me. Everything is now a "absolutely we can do that!" which becomes a "absolutely goatshaver will do this!" and i'm getting pretty over it.
|
|
#
?
Jan 8, 2016 16:52
|
|
|
SubjectVerbObject posted:You see this a lot in telecom stuff. Either ping will be turned off or it will drop every 3rd packet as some sort of rudimentary DDOS protection. I've been in telecom for a long time, this is completely new to me, granted, VOIP is a sideline for my normal security/network gig, but yeah. We have 3 other SIP gateways from this provider, none of them do this.
|
|
#
?
Jan 8, 2016 18:35
|
|
|
I'm working with a very large bank back-end vendor and their IT talent is absolutely horrendous. They have grown extremely quickly and acquired new services that they have no idea how to support, to which their solution was to hire anybody that could show up on time. We're moving from a dedicated T1 circuit to MPLS and good god these guys don't understand basic networking. We're on failure #2 to which they've admitted fault both times. Now a critical FTP service has stopped and this gem came from one of their system support guys: "FTP is not compatible with MPLS, so we did not move that functionality during yesterday's call. It seems that maybe there was a change at your end that inadvertently swung this traffic onto the new circuit. Could be a change on your mainframe, your firewall, we're not sure."  so many things to be pissed off at here.1) Review your OSI layers on FTP and MPLS and get back to me. 2) The change made was pointing our static route for your entire subnet from the old router to the new one. This means all services, including FTP, are routed that way. No, we cannot route FTP separately from other services going to the same IP range. We will have to create routes AROUND the one IP you want us to pass over the old connectivity, and gently caress you for that. 3) My motherfucking MainFrame? Is this a bad 1995 sci-fi movie??
|
|
#
?
Jan 8, 2016 18:38
|
|
|
Judge Schnoopy posted:I'm working with a very large bank back-end vendor and their IT talent is absolutely horrendous. They have grown extremely quickly and acquired new services that they have no idea how to support, to which their solution was to hire anybody that could show up on time. You might want to make sure the tape head on your reel-drive isn't gummed up.
|
|
#
?
Jan 8, 2016 18:41
|
|
|
DigitalMocking posted:I've been in telecom for a long time, this is completely new to me, granted, VOIP is a sideline for my normal security/network gig, but yeah. To be sure, it is usually devices with DSP resources that to VOIP processing. Given the real time nature of RTP, they don't like having to handle ping. The Oracle SBC is the old Acme Packet SBC, yes?
|
|
#
?
Jan 8, 2016 19:11
|
|
|
Judge Schnoopy posted:We're moving from a dedicated T1 circuit to MPLS and good god these guys don't understand basic networking. We're on failure #2 to which they've admitted fault both times. Now a critical FTP service has stopped and this gem came from one of their system support guys: Reuters till last year was using VAX hardware to communicate to NYSE and NASDAQ. Most circuits to clients used to be MPLS but are being replaced with various private lines such as Light Path and Verizon.
|
|
#
?
Jan 8, 2016 19:22
|
|
|
SubjectVerbObject posted:To be sure, it is usually devices with DSP resources that to VOIP processing. Given the real time nature of RTP, they don't like having to handle ping. The Oracle SBC is the old Acme Packet SBC, yes? I guess thinking about it, it does make sense now in a weird way. And yeah, Oracle bought them and rebranded the SBC
|
|
#
?
Jan 8, 2016 19:26
|
|
|
DigitalMocking posted:poo poo pissing me off: Bullshit replies to a trouble ticket. Oregon? I'm not a fan of Integra by any means, but the previous guy was. All of our services are through them and I've had a dozen outages over the last month. Thankfully it was mainly over the holidays when no one was here, but still. I'm looking to move away from them as soon as our contracts end.
|
|
#
?
Jan 8, 2016 19:55
|
|
|
DigitalMocking posted:poo poo pissing me off: Bullshit replies to a trouble ticket. Technically... ICMP is by definition low priority optional traffic and a device link that is oversaturated should drop ICMP traffic first (which is why TCP ECN is in the IP header), so it's a possibly bad way to check the connectivity of an overloaded device or link. But as with many subtle truths about technology, people parrot it incorrectly and build cargo cults around it like this one. Plus, it's Oracle, who would prefer you treat it as a magical block box and put in their EULA that you're not allowed to check it for bugs.
|
|
#
?
Jan 8, 2016 20:10
|
|
|
 Well... OK then.
|
|
#
?
Jan 8, 2016 20:26
|
|
|
They're supposed to say "differently abled" now anyway.
|
|
#
?
Jan 8, 2016 20:41
|
|
|
stevewm posted:
See you in 45 years wait, is that like 5 minutes from now but if the iPad thought it was Linux time 0? I think the clock is set wrong after being locked out. pixaal fucked around with this message at 20:53 on Jan 8, 2016 |
|
#
?
Jan 8, 2016 20:50
|
|
|
the spyder posted:Oregon? Yep, we're with integra for almost all of our services. They've been pretty good tbh, but we're just transitioning to SIP and seeing some hiccups.
|
|
#
?
Jan 8, 2016 20:54
|
|
|
keseph posted:Technically... ICMP is by definition low priority optional traffic and a device link that is oversaturated should drop ICMP traffic first (which is why TCP ECN is in the IP header), so it's a possibly bad way to check the connectivity of an overloaded device or link. But as with many subtle truths about technology, people parrot it incorrectly and build cargo cults around it like this one. Plus, it's Oracle, who would prefer you treat it as a magical block box and put in their EULA that you're not allowed to check it for bugs. Oh, that's true, but when I can ping one hop from the gateway with less than 4ms jitter, 0 loss, and 45ms trip time, I was freaked out to see 1500ms and 25% packet loss to the gateway itself. I still don't understand why the other gateways we use through Integra don't do this, but maybe they aren't Oracle.
|
|
#
?
Jan 8, 2016 20:56
|
|
|
pixaal posted:See you in 45 years Likely... this iPad was ripped from a retail display kiosk and has been sitting completely dead in a closet for some time.
|
|
#
?
Jan 8, 2016 21:08
|
|
|
DigitalMocking posted:Oh, that's true, but when I can ping one hop from the gateway with less than 4ms jitter, 0 loss, and 45ms trip time, I was freaked out to see 1500ms and 25% packet loss to the gateway itself. For one of the main carriers we use they explicitly have a separate device on their network to ping to/trace to, if you can reach that you can reach everything else. It's nice and has helped track down issues easily. Other carriers will easily ignore ping/trace issues and then turn around and ping your own public IP and say you have network issues.
|
|
#
?
Jan 8, 2016 23:20
|
|
|
stevewm posted:
There was a bug in older versions of IOS where repeatedly entering the wrong password would add 5 minutes to the lockout cumulatively. You'd have to be dedicated to get the timer up to 45 years, but I guess bored kids in a Verizon or whatever will try anything. You'll have to sync or wipe: https://support.apple.com/en-us/HT204306
|
|
#
?
Jan 8, 2016 23:53
|
|
|
Judge Schnoopy posted:3) My motherfucking MainFrame? Is this a bad 1995 sci-fi movie??
|
|
#
?
Jan 9, 2016 01:03
|
|
|
stevewm posted:
If you're willing to spend ~$30 for that iPad to work, plug it into a USB hub via the camera connector (so that it's the host), and then plug a USB Ethernet adapter into that hub. The iPad will go get the time via NTP and probably unlock.
|
|
#
?
Jan 9, 2016 01:09
|
|
|
Aunt Beth posted:Banking, insurance, and government is still dominated by mainframe computing. There's a reason why the platform still makes money hand over fist for IBM: World-class reliability, world-class performance, and complete backwards compatibility to the System/360 released in 1965. Big, slow, risk averse companies love the fact that they ironed all the bugs out of their core processes in 1978 and haven't had to modify since them. Wouldn't you too, when you're a house handling billions of transactions a day worth tens of billions of dollars? This. Mainframes are still incredibly common in business. If you haven't run into them good for you, I guess. If you feel like learning COBOL you can make good money. Lots of companies have essential processes written in COBOL, but everyone who knows it is in management or is retiring.
|
|
#
?
Jan 9, 2016 12:47
|
|
|
Kirios posted:Alright network dudes, what are you supposed to do when your boss is someone that worked at Cisco for 10+ years as a network engineer, acts like he knows everything, but actually knows very little and is full of terrible ideas? Meanwhile, you have to figure out how to implement said ideas and make them work while he sits at his desk, complains about being tired (you could try, oh I don't know, doing some work?) and takes all of the credit. Oh golly, does he also start every response to clarification questions with 'but it's so simple??'
|
|
#
?
Jan 9, 2016 13:40
|
|
|
Ynglaur posted:This. Mainframes are still incredibly common in business. If you haven't run into them good for you, I guess. If you feel like learning COBOL you can make good money. Lots of companies have essential processes written in COBOL, but everyone who knows it is in management or is retiring. I feel like it's worth pointing out that COBOL hasn't stayed the same language since 1965 or anything, and may of the programs written in it haven't either. It's also still taught a lot more than people think, even though it's nowhere near the level Java or any other popular language gets taught at.
|
|
#
?
Jan 9, 2016 16:19
|
|
|
Aunt Beth posted:Banking, insurance, and government is still dominated by mainframe computing. There's a reason why the platform still makes money hand over fist for IBM: World-class reliability, world-class performance, and complete backwards compatibility to the System/360 released in 1965. Big, slow, risk averse companies love the fact that they ironed all the bugs out of their core processes in 1978 and haven't had to modify since them. Wouldn't you too, when you're a house handling billions of transactions a day worth tens of billions of dollars? I used to work in a bank, writing COBOL and SAS and doing hideous things to z/OS job control language to make everything work. Goddamnit, I thought I'd drunk those memories away.
|
|
#
?
Jan 9, 2016 17:06
|
|
|
Ynglaur posted:This. Mainframes are still incredibly common in business. If you haven't run into them good for you, I guess. If you feel like learning COBOL you can make good money. Lots of companies have essential processes written in COBOL, but everyone who knows it is in management or is retiring. Reuters saved nearly $1 million per month turning off their VAX hardware, amazing savings. But then the HQ is in Times Square and the floor rent is $2 million per floor per year so 
|
|
#
?
Jan 9, 2016 17:23
|
|
|
Is COBAL really that bad? I wouldn't mind a chill job sitting in a dark basement somewhere programming.
|
|
#
?
Jan 9, 2016 21:03
|
|
|
Danith posted:Is COBAL really that bad? I wouldn't mind a chill job sitting in a dark basement somewhere programming. It really isn't, if you were to write a brand new program using modern COBOL. But if you're going into something that requires COBOL, there tends to be a bunch of code nearly as old as your dad that you're also responsible for, and the guy who wrote it originally died in 1987! So you're gonna need to spend a lot of time figuring out what the existing stuff does before you can coast. fishmech fucked around with this message at 21:34 on Jan 9, 2016 |
|
#
?
Jan 9, 2016 21:30
|
|
|
How hard is it to learn? Are they any good virtual environments on Windows or Linux where I could teach myself? I always thought having that skill and being able to travel would be a good backup job.
|
|
#
?
Jan 9, 2016 22:11
|
|
|
Danith posted:Is COBAL really that bad? I wouldn't mind a chill job sitting in a dark basement somewhere programming. Here's the Wikipedia section on COBOL's hello world for a quick glimpse at syntax. Aunt Beth fucked around with this message at 23:17 on Jan 9, 2016 |
|
#
?
Jan 9, 2016 23:14
|
|
|
Aunt Beth posted:Syntactically it's not bad at all, being the COmmon Business-Oriented Language, it was originally designed to be English-based and easily understood. I don't have any familiarity with SAS. JCL (Job Control Language) is a little dense, but roughly analagous to scripting in the open systems world. It's where you list out steps of things to run in a given job- feed program X dataset Y, output to printer Z, that sort of thing. whhhhhhhhhhhhhhhhhhhhhh
|
|
#
?
Jan 10, 2016 05:25
|
|
|
fishmech posted:It really isn't, if you were to write a brand new program using modern COBOL. But if you're going into something that requires COBOL, there tends to be a bunch of code nearly as old as your dad that you're also responsible for, and the guy who wrote it originally died in 1987! I'd be shocked if anyone developed anything new that was written in COBOL/RPG or any legacy language. Most legacy development will be spent modifying the existing code base for improvements and having it interact with modern IT Infrastructure. Remember, this stuff has been around since the 60s and it still runs and well in TYOOL 2016. Some organizations have conversion projects but it's analogous to moving mount Everest. It's massive undertaking for any company, extraordinary expensive, a risk to the business and often easier to just integrate.
|
|
#
?
Jan 10, 2016 05:57
|
|
|
Aunt Beth posted:Banking, insurance, and government is still dominated by mainframe computing. There's a reason why the platform still makes money hand over fist for IBM: World-class reliability, world-class performance, and complete backwards compatibility to the System/360 released in 1965. Big, slow, risk averse companies love the fact that they ironed all the bugs out of their core processes in 1978 and haven't had to modify since them. Wouldn't you too, when you're a house handling billions of transactions a day worth tens of billions of dollars? Isn't there something unique with the IBMi Hypervisor? I swear I remember reading a great article that went over virt CVEs and it went something like VMware 200, Hyper-V 300, Xen 200 and IBMi with only 2. That's extraordinarily impressive.
|
|
#
?
Jan 10, 2016 06:01
|
|
|
Tab8715 posted:Isn't there something unique with the IBMi Hypervisor? I swear I remember reading a great article that went over virt CVEs and it went something like VMware 200, Hyper-V 300, Xen 200 and IBMi with only 2. First, it's tied specifically to one closed platform (IBM Power) and not an open one (x86), so there's some security by obscurity there. Second, PHYP itself is less often manipulated directly by a customer in the way that any of the Intel hypervisors are. Partitions are managed by the HMC through the service processor network, which very rarely is on a LAN segment that carries anything but HMC traffic, so that attack vector is minimized as long as your HMC is patched. Third, IBM has built its hypervisor from the ground up solely as a hypervisor, with security in mind, whereas the others have been bolted onto other operating systems. If you're not talking about PHYP you're talking about PowerVM (formerly Advanced Power Virtualization) which is a product I know less about.
|
|
#
?
Jan 10, 2016 06:42
|
|
|
Tab8715 posted:Isn't there something unique with the IBMi Hypervisor? I swear I remember reading a great article that went over virt CVEs and it went something like VMware 200, Hyper-V 300, Xen 200 and IBMi with only 2. https://www.youtube.com/watch?v=Xfl4spvM5DI
|
|
#
?
Jan 10, 2016 07:06
|
|
|
I'd like to take a moment to appreciate all the cool people trapped in helldesk. 12 minutes on the phone (discounting hold but y'know) just took my internet from "we think you have a broken line and will have to come out in a week's time" to "poo poo, your speeds are faster than I thought we supported".
|
|
#
?
Jan 10, 2016 07:51
|
|
|
BurgerQuest posted:Oh golly, does he also start every response to clarification questions with 'but it's so simple??' Are you trying to trigger me because the answer yes of course he does. Why is it so simple? gently caress if he knows, but it is!
|
|
#
?
Jan 10, 2016 08:53
|
|
|
hoju22 posted:There was a bug in older versions of IOS where repeatedly entering the wrong password would add 5 minutes to the lockout cumulatively. You'd have to be dedicated to get the timer up to 45 years, but I guess bored kids in a Verizon or whatever will try anything. Nah 45 years is exactly 1970 to now (well last year). Its lost the ipad equivalent of the cmos clock and thinks its 1st January 1970, time_t of 0.
|
|
#
?
Jan 10, 2016 12:52
|
|
|
Tab8715 posted:I'd be shocked if anyone developed anything new that was written in COBOL/RPG or any legacy language. Most legacy development will be spent modifying the existing code base for improvements and having it interact with modern IT Infrastructure. I have a friend of the family whose job is maintaining stuff on the mainframes for a bank, and they do write new programs rather than just modifying existing ones. Of course, the company's practices require the new stuff to also be in COBOL (but it's the modern version of COBOL with object oriented stuff and the like), and the new programs don't get to go live until they've been tested for months or years.
|
|
#
?
Jan 10, 2016 17:06
|
|
|
pixaal posted:Printers effect production! They are the most important thing! I thought at first that it was because they were printing big, glossy brochures from our suppliers. But now we also have the issue with an excel that's about 360 KB, and a one-page color scan of an agenda sheet (just scanned the first thing that was at hand). The scan takes about 5 minutes to print on the Ricoh. It prints in about 10 seconds on a B/W brother.
|
|
#
?
Jan 11, 2016 11:18
|
|
|
|
| # ? Jun 8, 2024 14:25 |
|
|
BaseballPCHiker posted:I vaguely remember dealing with something similar in the past. I'll double check my notes and see if I can find the fix but I'm pretty sure it involved restarting the print spooler service on the computer that was slow to print. Thanks, I'll take any help I can get on this! poo poo's pissing me off, yo!
|
|
#
?
Jan 11, 2016 11:19
|
|