Options

1/256ths of a second (STP Timers)

showintshowint Member Posts: 68 ■■□□□□□□□□
Hello,

I'm considering BPDU Message Content. It's written in the table that Message age/ Maximum age/ Hello time/ Forward delay is 1/256ths of a second, with no further explanation. What does it (1/256ths of a second) exactly mean?

Thank you.

Comments

  • Options
    APAAPA Member Posts: 959
    Way out of scope for CCNA\CCENT but it is the format that the timer values are encoded into the BPDU for transmission.

    802.1D IEEE standard states:

    =====

    9.2.8 Encoding of Timer Values
    Timer Values shall be encoded in two octets, taken to represent an unsigned binary number multiplied by a unit of time of 1/256 of a second. This permits times in the range 0 to, but not including, 256 s to be represented.

    =====

    Basically what happens is we have two octets to represent timer values.

    Two octets == IIIIIIIIIIIIIIII = 16 bits

    If I turned on all those bits it would be 2^16 = 65536 (we count from 0 so basically that is 65535).

    1/256 of a second = 0.00390625

    65535 * 0.0390625 = 255.99609375

    The timer value range that can be represented in the two octets are from 0 - 255 (255.99609375)

    Hope that helps. I find it unlikely the CCNA\CCENT would test you on such fluff though..... nice to know but definitely not something you would ever need to call upon.

    CCNA | CCNA:Security | CCNP | CCIP
    JNCIA:JUNOS | JNCIA:EX | JNCIS:ENT | JNCIS:SEC
    JNCIS:SP | JNCIP:SP
  • Options
    PristonPriston Member Posts: 999 ■■■■□□□□□□
    So if the timer is 2 seconds its 00000010 00000000?
    A.A.S. in Networking Technologies
    A+, Network+, CCNA
  • Options
    showintshowint Member Posts: 68 ■■□□□□□□□□
    Priston, no, it is not.

    1 second = 1000 msec

    1/256 = 1000/256 ≈ 3,9 msec

    2 seconds = 2.000 msec

    2000 msec/3,9 msec ≈ 512

    that is, it will be 0000 0010 0000 0000
  • Options
    showintshowint Member Posts: 68 ■■□□□□□□□□
    APA, great explanation! Thank you!
  • Options
    APAAPA Member Posts: 959
    No problems.

    and you have it correct... 2 bytes considered as whole... 2 would be represented as 0000000000000010

    CCNA | CCNA:Security | CCNP | CCIP
    JNCIA:JUNOS | JNCIA:EX | JNCIS:ENT | JNCIS:SEC
    JNCIS:SP | JNCIP:SP
  • Options
    PristonPriston Member Posts: 999 ■■■■□□□□□□
    showint wrote: »
    Priston, no, it is not.

    1 second = 1000 msec

    1/256 = 1000/256 ≈ 3,9 msec

    2 seconds = 2.000 msec

    2000 msec/3,9 msec ≈ 512

    that is, it will be 0000 0010 0000 0000
    I'm confused, you just solved for 2 seconds and got the same answer as me.
    A.A.S. in Networking Technologies
    A+, Network+, CCNA
  • Options
    showintshowint Member Posts: 68 ■■□□□□□□□□
    Oh sorry. First I calculated incorrectly and thereby confused even APA. His answer was very detailed! super!

    You have it correct. I checked it in Wireshark.

    Hello Time:

    2 seconds = 0x0200, that is, 0000 0010 0000 0000 in binary

    Max age:

    1 second = 1000 msec
    1 sec/256 = 1000 msec/256 = 3.90625 msec
    20 seconds = 20 * 1000 = 20 000 msec
    20 000 msec/3,90625 msec= 5120
    5120 is 0001 0100 0000 0000 in binary

    If we compare it with WireShark's input, 0x1400 (which is 0001 0100 0000 000 in binary), it will match our calculated value.

    Forward Delay:

    15 seconds = 15 * 1000 = 15 000 msec
    15 000 msec/ 3.90625 msec = 3840
    3840 is 0000 1111 0000 0000 in binary

    Wireshark's input 0x0f00 (which is 0000 1111 0000 0000 in binary) matches as well, which in turn proves that we got it right.

    P.S. Now I'd like to learn how root path cost fits in 4 bytes)
  • Options
    APAAPA Member Posts: 959
    Yup, ignore my last post.

    I explained it well the first time..... then failed to double check your answer.

    Glad to be of some help ;)

    CCNA | CCNA:Security | CCNP | CCIP
    JNCIA:JUNOS | JNCIA:EX | JNCIS:ENT | JNCIS:SEC
    JNCIS:SP | JNCIP:SP
Sign In or Register to comment.