|
Frag Viper posted:Ok, but how do I figure out the possible combos? Thats the ONE thing driving me crazy right now and I want to learn. It's been a while since I've had to explain this so someone correct me if im wrong  If you take the subnet mask and turned it into binary it would be 11111111.11111111.11111111.11000000 Take the network and just lay it on top like the slut it is 11000000.10101000.00000100.00000000 11111111.11111111.11111111.11000000 The last dotted quad has 2 bits as a part of the network mask so 2^2 is 4 which are 00, 01, 10, and 11. So you have four networks. You get their boundaries by just doing the math: 00 = .0 01 = .64 10 = .128 11 = .192 Have you touched v6 yet? 
|
#
¿
Jan 23, 2013 05:24
|
|
|
|
| # ¿ Apr 23, 2024 15:22 |
|
|
CrazyLittle posted:We're not going to see IPv6 exhuastion... we're just going to see quadrillions of addresses wasted because entire /48 subnets will be relegated down to /126's in pager code. For example, Comcast's anycast DNSv6 IP: 2001:558:FEED::1 Yeah we have I think a /28 from ARIN and a /46 from APNIC and will probably get a shitton from RIPE. I'm swimming in v6 space. It makes it actually interesting to do network architecture at least. Only thing is that a /48 right now is equal to a /24 in v4 as far as the internet. The best waste is a /64 for a point to point link instead of a /127 psydude posted:Not cert related, but just out of curiosity given the discussion at hand: do you guys plan on still using unique local/private addresses for your networks once you start rolling out IPv6? I can see where it would have some definite security advantages over just using the global unicast address. There are philosophical arguments that will state each device on your network should have a globally unique address while others will want to stick to how they've been doing networking and use private addressing. Unique addressing would be easier and I don't know what security benefits you would gain from private addressing. doomisland fucked around with this message at 20:28 on Jan 23, 2013 |
|
#
¿
Jan 23, 2013 20:22
|
|
|
DropsySufferer posted:Any good estimate on how long IPv4 will remain the industry standard before IPv6 becomes the new standard? I've heard IPv4 will be the standard for at least another five years or more. Well, technically speaking the RIR's can get no more v4 address space. APNIC is on their last /8 which means they're on stage 3 of their IPv4 exhaustion plan. RIPE is on their last /8. ARIN has a little over 2 /8's though at least. As far as industry standard all the ISPs at least are probably 1-3 years out from full v6 and if they're not they're poo poo. Comcast is the most ahead from what I can tell. They've been upgrading their CMTS equipment last I knew to handle IPv6. They also have a handy blog http://www.comcast6.net/ Each of of the RIRs have plans for exhaustion and such on their sites: http://www.ripe.net/internet-coordination/ipv4-exhaustion https://www.arin.net/resources/request/ipv4_countdown.html http://www.apnic.net/community/ipv4-exhaustion http://www.lacnic.net/en/web/lacnic/reporte-direcciones-ipv4 http://afrinic.net/en/services/statistics/ipv4-exhaustion CrazyLittle posted:As long as it takes for people to ditch their old routers and firewalls. Five years is a modest estimate. I think it will be in wide use in five years, but probably won't overtake IPv4 universally for seven or more years. There are people who are still using IPX/SPX for their LAN protocols. You're wasting two IPs using /126 :p doomisland fucked around with this message at 20:50 on Jan 23, 2013 |
|
#
¿
Jan 23, 2013 20:48
|
|
|
Can't believe people use classfull addressing still. Just to let know if you mention classes around a bunch of network engineers they'll probably laugh at you. What I'm saying is make sure you understand CIDR
|
|
#
¿
Jan 26, 2013 15:38
|
|
|
hooah posted:Wait, I thought I was using CIDR, since I know about the slash notation and how it applies to the masks (the number of the slash is the number of on bits). That still doesn't help me when I have to subnet in the third octet. I was more commenting on people still saying 'Class A/B/C' which no one uses anymore and haven't forever. It's only use is for understanding legacy networking. Keep on using /XX for subnets.
|
|
#
¿
Jan 26, 2013 19:21
|
|
|
hooah posted:That's all well and good, but it doesn't help me understand how to subnet an address that has host bits in the third octet. You should write everything out in binary to figure it out to understand it. You have your network mask and host mask. Your block range is your host mask digits being all 0's to being all 1's. As mentioned before you can figure out your ranges from the highest bit in your host mask. So if you have 172.16.0.0/19 your subnet is the following 11111111.11111111.11100000.0000000 The highest bit in the host mask is in the third octet and is the 5th (2^5 is 32) bit in it. Your netmask has 3 bits in the network mask in the third octet so you have 2^3 networks (which equals 8). So, now you have to figure out your boundaries which is just math. 172.16.0.0 - 172.16.31.255 | 10101100.00010000.00000000.00000000 - 10101100.00010000.00011111.11111111 172.16.32.0 - 172.16.63.255 | 10101100.00010000.00100000.00000000 - 10101100.00010000.00111111.11111111 172.16.64.0 - 172.16.95.255 | 10101100.00010000.01000000.00000000 - 10101100.00010000.01011111.11111111 172.16.96.0 - 172.16.127.255 | 10101100.00010000.01100000.00000000 - 10101100.00010000.01111111.11111111 172.16.128.0 - 172.16.159.255| 10101100.00010000.10000000.00000000 - 10101100.00010000.10011111.11111111 172.16.160.0 - 172.16.191.255| 10101100.00010000.10100000.00000000 - 10101100.00010000.10111111.11111111 172.16.192.0 - 172.16.223.255| 10101100.00010000.11000000.00000000 - 10101100.00010000.11011111.11111111 172.16.224.0 - 172.16.255.255| 10101100.00010000.11100000.00000000 - 10101100.00010000.11111111.11111111
|
|
#
¿
Jan 26, 2013 20:44
|
|
|
I've heard folks say CCIE's are useless now since everyone has them and can just study the answers and not know how to do networking. Crazy world out there.
|
|
#
¿
Jun 2, 2015 23:16
|
|
|
crunk dork posted:You could probably dump any written exam if you tried hard enough, not so sure about hands on labs though. You'd think the answers would be so complex at that level of knowledge that memorizing the commands necessary to complete and order of executing them would be harder than just learning the material itself. Yeah you would think but one example was them not knowing how IPv6 works and differences between frames and packets.
|
|
#
¿
Jun 2, 2015 23:27
|
|
|
MC Fruit Stripe posted:There's always going to be outliers and people to disagree with me, but I'll say this. I think if you can study test dumps in such a way that it allows you to bullshit your way to a CCIE, that without even having meant to, you've backdoored your way into being a pretty decent network administrator. My point being, even the worst CCIE in the world is probably, let's say, a strong CCNA. The general point they were trying to make is certs dont mean poo poo and they'll take dudes who they know are good with nothing to their name over someone with every CCIE Cisco offers. From what I've seen"good" in this context means someone who can pick stuff up quick and has the ability to then adapt and troubleshoot. Poking at MPLS for the first time to implementing it the next week etc etc doomisland fucked around with this message at 01:48 on Jun 3, 2015 |
|
#
¿
Jun 3, 2015 01:45
|
|
|
|
| # ¿ Apr 23, 2024 15:22 |
|
|
evol262 posted:Sounds like someone has a chip on his shoulder and/or is a bad interviewer. Again, certs are not a substitute for experience. You still need to interview candidates. Good CCIEs are great. Certs establish some familiarity with best practice even if braindumped, which "poking at MPLS" won't tell you. You shouldn't devalue them, but you shouldn't value them in a vacuum either. Does this make sense? The world is not black and white, and, as a generalization, CCIEs (and certs in general) are valuable No I get that it's just interesting seeing the two sides of this.
|
|
#
¿
Jun 3, 2015 04:06
|
|