Kitz Forum
Broadband Related => ADSL Issues => Topic started by: noddy on December 08, 2021, 04:22:51 PM
-
I wonder if some one could explain why there is such a difference between Max up ? down stream rate and actual rates ?
Max: Upstream rate = 779 Kbps, Downstream rate = 4548 Kbps
Bearer: 0, Upstream rate = 833 Kbps, Downstream rate = 3071 Kbps
cheers
-
The "Bearer: 0, . . . " line reports the current reality. The "Max: . . . " line reports a guess (or estimate) of what the modem (or modem/router) believes could be achieved "if all things were perfect" (whatever that five word phrase means).
In other words: Reality is "Upstream rate = 833 Kbps" whilst a guess (or estimate) is "Upstream rate = 779 Kbps".
-
Thanks for that , all clear now :)
-
The "Bearer: 0, . . . " line reports the current reality. The "Max: . . . " line reports a guess (or estimate) of what the modem (or modem/router) believes could be achieved "if all things were perfect" (whatever that five word phrase means).
In other words: Reality is "Upstream rate = 833 Kbps" whilst a guess (or estimate) is "Upstream rate = 779 Kbps".
Perfectly explained as always, Mr Cat - methinks noddy was after the reason behind the downstream value chasm, but of course your same logic applies. :)
-
Perfectly explained as always, Mr Cat - methinks noddy was after the reason behind the downstream value chasm, but of course your same logic applies. :)
you are quite right I was ;) but it answered both figures :)
-
I live in hope of the downstream figure becoming true one day :fingers:
-
There is just one other thing to mention . . .
Those values (on the "Bearer: 0, . . ." and the "Max: . . ." lines) are of the synchronisation speed of the link and not the end-user's throughput speed. The throughput speed is the speed that is achieved whilst using the service; for example, running a "speed test" or posting to this forum.
The throughput speed will always be less than the synchronisation speed due to the various protocol overheads.
-
well the IP profile of 2.71 makes sense with a max of 3071
-
<Nods.> A throughput speed of around 2.7 Mbps is about right with a synchronisation speed of around 3.0 Mbps.
(Our friend Weaver is quite adept at laying out all the arithmetic adjustments, and then doing the calculations, to produce the expected end result.)
-
Our friend Burrakucat mentions the difference between sync rate and IP PDU speed, or IP SDU speed. These numbers are in decreasing order. The difference is because of various overheads, headers, possibly trailers and also inefficiency of coding schemes that represent sequences of bytes by more longwinded sequences of bits. A PDU is a packet where the rate number quoted includes the cost of transmitting headers, trailers and all other overheads; it’s a larger number. An SDU is the payload/content of a packet, excluding headers and ignoring any encoding; it is the data passed down to a service to be sent or the received data passed up by a service. This is a smaller number. SDU rates are what users expect from services; PDU rates measure the link itself as opposed to the accidental payload throughput that some particular protocol achieves, and different protocols achieve different amounts of SDU (ie payload) throughput.
For example IPv4 is more efficient than IPv6 because IPv4 has a 20 byte header whereas IPv6 has (at least) a 40 byte header. When including the popular TCP protocol that means PDUs maybe 1500 bytes long and an IPv4+TCP SDU will be either 1500-(20+20)=1460 bytes long or 1500-(20+20+12) bytes long if TCP uses timestamps at a cost of an additional 12 bytes of headers. So the payload of IPv4+TCP (the SDU size) will be either 1460 or 1448 bytes of user data. For IPv6+TCP the SDU size would be 1500-(40+20) or 1500-(40+20+12) bytes.
VDSL2/FTTC for example has additional overheads that are not included in the above; these include the bloat due to encoding, ethernet headers and other headers. The encoding bloat reduces efficiency to either 96.69% / 97.69% or 91% depending on protocol settings in force which may vary over time. All the header overheads can iirc vary with the ISP.
ADSL/ADSL2/ADSL2+ has very inefficient overheads. My ADSL2 service has 32 bytes of headers and an encoding system that adds on a variable amount of bloat but in the case of a long packet the bloat expansion factor is 53:48; for a short packet the expansion factor can be 100% [!], and a complex formula is required to calculate the exact length. Including all inefficiencies, a 1500 byte payload ADSL+PPP+PPPoEoA PDU can have an efficiency of ~88.44% because 1500 bytes (SDU size) expands to 1696 bytes (PDU size).
-
Just to check my / our understanding: Are the "Max" values not expressed in exactly the same terms as the "Bearer" values ? That is, no protocol overheads are included/excluded. Is that what we’re saying?
-
Telnet data
adsl info --stats
adsl: ADSL driver and PHY status
Status: Showtime
Last Retrain Reason: 8000
Last initialization procedure status: 0
Max: Upstream rate = 779 Kbps, Downstream rate = 4520 Kbps
Bearer: 0, Upstream rate = 833 Kbps, Downstream rate = 3071 Kbps
Link Power State: L0
Mode: ADSL2 Annex A
TPS-TC: ATM Mode(0x0)
Trellis: U:ON /D:ON
Line Status: No Defect
Training Status: Showtime
Down Up
SNR (dB): 6.8 6.1
Attn(dB): 55.5 35.8
Pwr(dBm): 18.2 12.3
ADSL2 framing
Bearer 0
MSGc: 89 12
B: 10 94
M: 4 1
T: 6 1
R: 16 16
S: 0.4440 3.6098
L: 1081 246
D: 4 8
Counters
Bearer 0
SF: 41655628 961420
SFErr: 1246 180
RS: 1640959780 3544741
RSCorr: 455092 2975219
RSUnCorr: 11278 0
ReXmt: 100440 0
ReXmtCorr: 63145 0
ReXmtUnCorr: 11278 0
Bearer 0
HEC: 7914 457
OCD: 488 0
LCD: 488 0
Total Cells: 558289110 1316625203
Data Cells: 388930575 32494985
Drop Cells: 0
Bit Errors: 0 62082
ES: 760 282
SES: 30 0
UAS: 192 181
AS: 669929
Bearer 0
INP: 28.00 2.00
INPRein: 0.00 0.00
delay: 8 7
PER: 15.81 16.24
OR: 48.04 8.86
AgR: 3086.80 838.88
Bitswap: 32783/32821 48/48
Total time = 18 days 15 hours 34 min 33 sec
FEC: 5068000 3622550
CRC: 3493 386
ES: 760 282
SES: 30 0
UAS: 192 181
LOS: 1 0
LOF: 9 0
LOM: 0 0
Latest 15 minutes time = 4 min 33 sec
FEC: 2 1716
CRC: 0 0
ES: 0 0
SES: 0 0
UAS: 0 0
LOS: 0 0
LOF: 0 0
LOM: 0 0
Previous 15 minutes time = 15 min 0 sec
FEC: 27 5668
CRC: 0 0
ES: 0 0
SES: 0 0
UAS: 0 0
LOS: 0 0
LOF: 0 0
LOM: 0 0
Latest 1 day time = 15 hours 34 min 33 sec
FEC: 3055 284733
CRC: 102 15
ES: 30 11
SES: 0 0
UAS: 0 0
LOS: 0 0
LOF: 0 0
LOM: 0 0
Previous 1 day time = 24 hours 0 sec
FEC: 3872 377215
CRC: 86 41
ES: 36 31
SES: 0 0
UAS: 0 0
LOS: 0 0
LOF: 0 0
LOM: 0 0
Since Link time = 7 days 18 hours 5 min 27 sec
FEC: 455092 2975219
CRC: 1246 180
ES: 325 131
SES: 8 0
UAS: 0 0
LOS: 0 0
LOF: 0 0
LOM: 0 0
>
-
Just to check my / our understanding: Are the "Max" values not expressed in exactly the same terms as the "Bearer" values ? That is, no protocol overheads are included/excluded. Is that what we’re saying?
Yes, they are just referring to the circuit synchronisation speed.