#cisco
26 December 2007
Total 21 pages. You are browsing page 2/21.
First :: Prev :: [1] [2] [3] [4] [5] [...] :: Next :: Last
05:02 <****> but i'm sure the how much you can change it depends on the hardware
05:02 <****> so the packet isn't physically in the ring
05:03 <****> dude, how do you know this shit, seriously
05:03 <****> let me check this i was reading, 1 sek
05:03 <****> jhuebel: quite frankly i don't think your living situation has anything to do with P4S
05:04 <****> if you need the money.. study harder
05:04 <****> mate I agree with P4S and TK and all those braindump stuff killing the prestige of Certification
05:04 <****> rationalizing cheating is so unbecoming
05:04 <****> I find going into an interview having them on ur CV makes the employer think "oh here we go, lets drill this guy!"
05:05 <****> it's ok though
05:05 <****> cheating rationally, on the other hand...
05:05 <****> your CCNP cert now is less of a vaule than it was even 4 years ago
05:05 <****> yup
05:06 <****> at work I use to get heaps from others that it took me a year to do my CCNP
05:06 <****> some software queue commands automatically adjust the hardware queue size, and you can change the hardware queue with tx-ring-limit
05:06 <****> they would say...CCNP can take 1 month to do
05:06 <****> so tx-ring-limit changes the actual buffer
05:06 <****> the hardware buffer
05:07 <****> CCNP definitely can't be done in a month, unless that's the only thing you do all day, every day.
05:07 <****> Even then...
05:07 <****> It took me 16 months, mainly because of my work schedule.
05:07 <****> or TK and P4S
05:07 <****> cpu utilization needs to be accounted for, and will change when changing the hardware queue size
05:07 <****> jhuebel: i did it in about 2 month
05:08 <****> yup same and because I actually wanted to learn and understand it
05:08 <****> 2 months would be constant work everyday i think
05:08 <****> dwxreaper: no.. tx-ring-limit is the ring
05:08 <****> nemith: Awesome. Wish I could do that, but I was running my own company and such.
05:08 <****> like i said, it's just as pointers to the part of memory with the real packet
05:08 <****> So that was a 60-hour work week on its own.
05:08 <****> the packet really doesn't move though the rings, just pointers are moved around
05:08 <****> you can do ccnp in 1 month if you only are trying to pass the exams
05:09 <****> CCIP took me 4 months becauase of the work I was doing was the theory behind it all so I understood it straight up
05:09 <****> november2: well my boss wanted me to do it in a month
05:09 <****> so i was studying a lot
05:09 <****> nemith: memory on the interface? because this is saying that command changes the hardware queue size, and this is the ciscopress book
05:09 <****> or if you already know a lot
05:09 <****> lol
05:09 <****> yup
05:09 <****> dwxreaper: no it doesn't
05:09 <****> usually the packet isn't stored on the interface
05:10 <****> november2: I did a lot of simulators, etc.
05:11 <****> MPLS-TE: that would explain the nick
05:11 <****> "you can set the size of the hardware queue by using the tx-ring-limit command"
05:11 <****> so if that command doesn't change the software queue, what is referred to as the hardware queue
05:11 <****> hell the ccie just took me 6 months, but a bunch of hours
05:11 <****> CCNP is indeed cheapened a lot nowadays, even entry level networking jobs state CCNP is pre-requisite
05:12 <****> Loatheb: exactly, cause idiots can use p4s and tk to get it and then get the job and still no nothing
05:12 <****> hehehe
05:12 <****> nemith: yes it would
05:12 <****> ccnp is harder than it was before, much harder
05:12 <****> like the ccna
05:12 <****> I just did JNCIS-M last week
05:12 <****> i hate terms like "Well the CCNP has IS-IS on it.. or frame-relay and I don't need to learn that cause it's old technology that i will never use"
05:12 <****> The certificiation are to test you on what is deamed the level of a CCNP
05:12 <****> and in the new year will start preparation for CCIE SP written and then lab end of 08
05:12 <****> dwxreaper; its called the hardware queue because the software queue is the llq or cbwfq, but it may still not be a local l2 cache on the interface
05:13 <****> not what you think it should be
05:13 <****> then again, it may be on some hardware
05:13 <****> dwxreaper: which the actually cache on the interface (if any) wouldn't be software settable with a command
05:13 <****> MPLS-TE: Good luck! :-)
05:13 <****> thats should be clue enough
05:14 <****> course i am talking about software platforms
05:14 <****> which is what your book is refering too
05:14 <****> nemith: yeah because you can't reprogram that circuitry, i guess, i thought you could alter the way transitors move packets at that level
05:14 <****> i'm obviously a little confused :)
05:14 <****> dwxreaper: well it's possible, but not cost effective
05:15 <****> dwxreaper: there are took books that really explain how packets move though a cisco box
05:15 <****> the CEF book and the Router Archetecture books
05:15 <****> if you really want to know I recommend both of them
05:15 <****> and both are a quick read
05:15 <****> nemith: cool, i'd definitly like to read them, but i finally realized i have to get this np before march
05:15 <****> so it's learning what cisco tells me for 2 months :P
05:16 <****> dwxreaper; fpgas are what you'd be refering
05:16 <****> i'm not big on how auto-qos is such a huge part of the exam, and wirelss (because i don't have the means to learn it)
05:17 <****> they're fun to play with
05:17 <****> locuer: i've read a little about them
05:18 <****> i was worried about electrecuting myself fixing this power adapter yesterday :)
05:20 <****> where is the link linked list of pointers stored, the list which is the tx-ring
05:21 <****> i'm guessing it's pointers, because of the linked list contained the packets, it just wouldn't worked, pointers only reference the memory
05:21 <****> the whole packet doesn't need to be contained in the txring
05:21 <****> moving bits is expensive
05:21 <****> why move bits around in memory
05:22 <****> understanding pointers really needs a C background
05:22 <****> but it's just the memory location of the packet
05:22 <****> i know what pointers are, just not where the memory is
05:23 <****> i guess it's platform dependent
05:23 <****> so the same part of memory is referenced on the RX_ring on the ingress interface as it is on the TX_ring on th egress interface
05:23 <****> well again
05:23 <****> this only applies to software
05:23 <****> where there is shared memory
05:23 <****> omg, cisco needs to start revamping theirs cisco 800 series..where is the g.lite support?
05:24 <****> as soon as you have hardware in the mix you probably don't have a shared memory infastructure due to realability and speed
05:24 <****> nemith: so the memory that holds the linkedlist of pointers which is the tx or rx ring is in the shared memory
05:24 <****> or on the interface itself, depending on the platform
05:25 <****> guess i gotta read that book on cef and the router architecture
05:25 <****> yes
05:25 <****> thats a great idea
05:25 <****> Nitrate: 800's is for children
05:26 <****> there is about 3 different software platforms which all work just a little different
Total 21 pages. You are browsing page 2/21.
First :: Prev :: [1] [2] [3] [4] [5] [...] :: Next :: Last