mdm@cocktrice.UUCP (Mike Mitchell) (12/14/88)
I have recently added a Telebit TrailBlazer+ to my system, and I am wondering if anyone out there has the optimal setup parameters to make this beast scream with V/AT? I currently have the beast configured and operating with a couple of links, so it is operational. There were some changes that I made to the default factory configuration. These changes have come from various places on the net, however they are not specific to V/AT. So far I am averaging somewhere on the order of 675 chars/sec for file transfer (this blows away what I was doing), however I am wondering if there might be a nifty way to bump this up. Thanks for any pointers. -- Mike Mitchell BELL: (505) 471-7639 2020 Calle Lorca #43 ARPA: mdm@cocktrice.UUCP Santa Fe, NM 87505 UUCP: ...!uunet!dmk3b1!cocktrice!mdm
debra@alice.UUCP (Paul De Bra) (12/15/88)
In article <354@cocktrice.UUCP> mdm@cocktrice.UUCP (Mike Mitchell) writes: >I have recently added a Telebit TrailBlazer+ to my system, and I am >wondering if anyone out there has the optimal setup parameters to make >this beast scream with V/AT? >... >So far I am averaging somewhere on the order of 675 chars/sec for file >transfer (this blows away what I was doing), however I am wondering if >there might be a nifty way to bump this up. You don't specify what program you use for file-transfer, but in the case of uucp 675 chars/sec is pretty respectable, especially with V/AT (which is slow in general). I have used uucp on all sorts of lines, baud-rates and distances and you generally get at most 2/3 of the baud-rate (except at very low speeds when you can get more). The only way to increase speed beyond this is to use protocol "f", but not all uucp implementations support it, and it only gives you a 7-bit datapath. Paul. -- ------------------------------------------------------ |debra@research.att.com | uunet!research!debra | ------------------------------------------------------
rsj@wa4mei.UUCP (Randy Jarrett WA4MEI) (12/16/88)
In article <354@cocktrice.UUCP> mdm@cocktrice.UUCP (Mike Mitchell) writes: >I have recently added a Telebit TrailBlazer+ to my system, and I am >wondering if anyone out there has the optimal setup parameters to make >this beast scream with V/AT? > >I currently have the beast configured and operating with a couple of >links, so it is operational. There were some changes that I made to the >default factory configuration. These changes have come from various >places on the net, however they are not specific to V/AT. > >So far I am averaging somewhere on the order of 675 chars/sec for file >transfer (this blows away what I was doing), however I am wondering if >there might be a nifty way to bump this up. > >Thanks for any pointers. > > >-- >Mike Mitchell BELL: (505) 471-7639 >2020 Calle Lorca #43 ARPA: mdm@cocktrice.UUCP >Santa Fe, NM 87505 UUCP: ...!uunet!dmk3b1!cocktrice!mdm Below is the setup that I am using with my TB+ on V/AT 2.4. I am also using a version of uugetty that changes the value in S50 to 255 when I am calling a system at the higher speed. Also important is to set S92 to 1 which will cause the system to answer with the PEP tones last so as to not drive the 1200 and 2400 baud modems crazy. As far as transfer speed goes the second section below shows my system averages for two days. The sites kd4nc, gatech, and nanovx are running trailblazers. kd4nc is also a V/AT 2.4 system and nanovx is a Xenix system and as you can see the best transfers seem to be to nanovx. I'll leave any conculsions to you. I am currently running a Trailblazer and 1 hayes 2400 modems. My second Trailblazer is on the way and the hayes will be put on the shelf with its mate. I have a Connect Tech 8 port inteligent board and kd4nc is running their 4 port. The boards work real well and we havn't had any problems with them. E0 F1 M1 Q0 P V1 X1 Version BA4.00 S00=004 S01=000 S02=255 S03=013 S04=010 S05=008 S06=002 S07=040 S08=002 S09=006 S10=007 S11=070 S12=050 S45=000 S47=004 S48=000 S49=000 S50=000 S51=255 S52=002 S53=001 S54=002 S55=000 S56=017 S57=019 S58=002 S59=000 S60=000 S61=045 S62=003 S63=001 S64=000 S65=000 S66=000 S67=000 S68=002 S90=000 S91=000 S92=001 S95=002 S100=000 S101=000 S102=000 S104=000 S110=255 S111=255 S112=001 S121=000 Remote K-Bytes K-Bytes K-Bytes Hours Hours AvCPS AvCPS # # SiteName Recv Xmit Total Recv Xmit Recv Xmit Recv Xmit -------- --------- --------- --------- ------ ------ ----- ----- ---- ---- kd4nc 0.300 1726.845 1727.145 0.00 0.60 38 804 6 90 wgs6386 0.000 867.041 867.041 0.00 1.09 0 220 0 546 dscatl 0.879 1981.860 1982.739 0.00 3.95 89 139 4 106 emcard 0.313 1.086 1.399 0.00 0.00 83 242 2 4 gatech 1581.256 0.000 1581.256 0.95 0.00 464 0 68 0 nanovx 0.000 5744.270 5744.270 0.00 1.13 0 1417 0 272 bcs3b2 0.504 0.000 0.504 0.00 0.00 95 0 2 0 Remote K-Bytes K-Bytes K-Bytes Hours Hours AvCPS AvCPS # # SiteName Recv Xmit Total Recv Xmit Recv Xmit Recv Xmit -------- --------- --------- --------- ------ ------ ----- ----- ---- ---- kd4nc 0.100 2434.802 2434.902 0.00 1.06 22 636 2 112 gatech 3189.493 0.285 3189.778 1.91 0.00 464 252 142 2 emcard 0.557 0.287 0.844 0.00 0.00 119 282 2 2 nanovx 0.000 3382.175 3382.175 0.00 0.64 0 1457 0 151 dscatl 0.000 659.776 659.776 0.00 0.97 0 189 0 30 -- Randy Jarrett WA4MEI UUCP ...!gatech!wa4mei!rsj | US SNAIL: P.O. Box 941217 PHONE +1 404 493 9017 | Atlanta, GA 30341-0217
andrew@ramona.UU.NET (Andrew Ernest) (12/18/88)
From the register dump he included in his article, it appears Randy Jarrett is attempting to use RTS/CTS flow control (S58 = S68 = 2) between his computer and Trailblazer modem. I was under the impression that V/AT did not support this form of handshaking. I tried it with 2.3 and the serial driver seems to ignore RTS/CTS. Fortunately, uucp's g protocol does its own flow control. However, other applications would be much cleaner and simpler if V/AT supported RTS/CTS handshaking. Microport tech support (or anyone else who knows for sure), does V/AT support RTS/CTS flow control? If so, how does one enable it? I would love to be able to lock the interface speed and dispense with XON/XOFF flow control for non-uucp applications. Andrew Ernest (andrew@ramona.uu.net)
rsj@wa4mei.UUCP (Randy Jarrett WA4MEI) (12/19/88)
In article <412@ramona.UU.NET> andrew@ramona.UU.NET (Andrew Ernest) writes: >From the register dump he included in his article, it appears Randy >Jarrett is attempting to use RTS/CTS flow control (S58 = S68 = 2) >between his computer and Trailblazer modem. I was under the >impression that V/AT did not support this form of handshaking. I >tried it with 2.3 and the serial driver seems to ignore RTS/CTS. I don't know if V/AT supports RTS/CTS or not. As I mentioned, I am using the Connect Tech 8 port intelligent serial board which as far as I know, supports the handshaking. I am running 1 TB+, 1 Hayes 2400 (soon to be TB+), and 2 19,200 serial ports on it so far and have not had any problems at all. There are some others in the area that are also using the CT board and are happy with it. -- Randy Jarrett WA4MEI UUCP ...!gatech!wa4mei!rsj | US SNAIL: P.O. Box 941217 PHONE +1 404 493 9017 | Atlanta, GA 30341-0217
jiii@visdc.UUCP (John E Van Deusen III) (12/31/88)
From your list of register contents, the setting S51=255 (automatic speed select) defaults to 9,600 baud. There is a poorly-documented setting of S51=254 that defaults to 19,200 baud. The TB+ really needs to be fed at 19,200 because the theoretical data rate is higher than 9,600 baud. The higher rate gives the processor more time to do compression and error correction too. A 9,600 baud interface speed at the receiving end is less critical because the processing has already been done. Some systems don't like unsolicited input at 19,200. A backbone site like uunet, that transmits at 19,200 baud (S51=254), can very effectively send news to a site running a TB+ at 9,600 (S51=255). Since you have an I/O board with an effective flow control protocol, I can't see why you wouldn't want to simplify your life and set S51=5, 19,200 baud interface speed at all transmission modes. You can feed the TB+ at 19,200 and the modem with take care of the rest, even talking to a Bell 103 with no modem control.
debra@alice.UUCP (Paul De Bra) (01/01/89)
In article <252@visdc.UUCP> jiii@visdc.UUCP (John E Van Deusen III) writes: >... >Since you have an I/O board with an effective flow control protocol, >I can't see why you wouldn't want to simplify your life and set >S51=5, 19,200 baud interface speed at all transmission modes. You >can feed the TB+ at 19,200 and the modem with take care of the rest, >even talking to a Bell 103 with no modem control. Sure uucp can talk to a Bell 103, but when you log on from a lower-speed modem and try to work interactively, the buffers will overflow. When you try an ls -l in a large directory the output may be garbled after some number of lines because of buffer overflow in the modem. Only programs that use relatively small packets will work well with speed-mismatches. Paul. -- ------------------------------------------------------ |debra@research.att.com | uunet!research!debra | ------------------------------------------------------