The timing diagram showm below depicts host A\'s use of a persistent noa-pipelin

Question

The timing diagram showm below depicts host A's use of a persistent noa-pipelined ITTTP connection to retrieve a base page and two embedded objects from host B. host A bost B t secs Real tme 4 secs xind tizne 4 secs obj 2 amt timo 4 secs xunt time time ume (4) 6. How much time would use of s persistent, pipelined HTTP connection save? (a) 1 scc. (b) 2 secs. (c) 3 secs. (d) 4 secs. ( (e) 5 secs. () 6 secB. (4) 7. How much time would use of a non-persistent, non-pipelined HTTP connectiou lose? (a) 2 secs. (b) 4 secs. (e) 5 secs. (d) 8 secs. (e) 10 secs. (0) 12 secs (5) 8. When TCP flows, TCP1 and TCP2, desiring rates of 0.4 Mbps and 0.3 Mbps share a 1 Mbps link witb a UDP flow desiring rate 0.4 Mbps, the eventual average rates of operation of the three lows will be RA, RB, and Re DESIRED RATE EVENTUAL RATE Shared 1 Mbpe link RA TCP1 0.4 Mbps TCP2 0.3 Mbps UDP 0.4 Mbps Ro HINT: Consider separately how the UDP and TCP flows react to congestion. The eventual rates RA, RB, and Rg are (a) 0.2 0.2 0.6 (b) 0.3 0.3 0.4 (c) 1/3 1/3 1/3 (d) 0.3 0.1 0.6 (e) 0.4 0.2 0.4 (f) 0.4 0.3 0.3

Explanation / Answer

In HTTP there are 2 types, persistent and non persistent HTTP.

Either the connection is persistent or non persistent at the initial

stage we need 2 RTT one for TCP connection and other for reference page (Obj 1).

Now if it is persistent HTTP then for every object 1 HTTP

connection (1 RTT) is needed on the other hand

if it is non persistent HTTP for every object 1 TCP connection (1 RTT)

and 1 HTTP connection (1 RTT) a total of 2 RTT requried.

In other classification we have 2 types of HTTP one is

Pipelined HTTP and other is non pipelined HTTP.

In non pipelined HTTP for every object 1 HTTP

connection (1 RTT) is needed on the other hand

In pipelined HTTP for all objects 1 HTTP connection (1 RTT) is needed.

6)

     Given HTTP is persistent non pipelined and it takes 2 RTT for

     TCP and reference page(obj 1) and 2 RTT for obj 2 and obj 3.

     Total 4 RTT = 4 * 4 = 16 secs.

      Now if we use persistent pipelined HTTP then it takes 2 RTT for

      TCP and reference page(obj 1) and 1 RTT for obj 2 and obj 3.

      Total 3 RTT = 3 *4 = 12 secs.

      Hence we saved 4 secs by using pipelined HTTP.

7)

      Given HTTP is persistent non pipelined and it takes 2 RTT for

     TCP and reference page(obj 1) and 2 RTT for obj 2 and obj 3.

     Total 4 RTT = 4 * 4 = 16 secs.

      Now if we use non persistent non pipelined HTTP then it takes

      2 RTT for TCP and reference page(obj 1) and 2 RTT for obj 2 (TCP + obj2)

       and 2 RTT for obj 3 (TCP + obj3).

       Hence a total of 6 RTT are requried which is = 6 * 4 secs = 24 secs

       Hence if we use non persistence non pipelined connection then we loss 24 - 16 = 8 secs.

8)

       In flow control all TCP flows will have same bandwidth and

      UDP will use as much it wants.

      Hence UDP rate RC = 0.4 Mbps

      The remaining 0.6 Mbps will be shared by 2 TCP flows equally

      Hence RA = 0.3 Mbps RB = 0.3 Mbps

Related Questions

Navigate

• Browse (All)
• Browse T
• Subjects

---

---

Integrity-first tutoring: explanations and feedback only — we do not complete graded work. Learn more.