- WLUG - wlug.mailman3.com

Re: poor PCMCIA network performance

by Mark J. Dulcey

Robert L Krawitz wrote: > I use two network cards in my Dell Inspiron 8000 (PIII-700): a 3Com > 3CCFE574BT and a Linksys wireless card. With the 3Com card (which is > a 10/100) I never get better than 1.4 MB/sec (about 12 Mb/sec), even > when the card is in 100 Mb mode (at some point in the past the limit > was about 2.5 MB/sec). At that rate, top shows my CPU utilization > pegged at about 100%, even doing a simple ftp. I know the card > doesn't do DMA, but that's rather slow even for a PIO transfer. These > cards have been around for ages; I don't think people would have been > too satisfied with 400 KB/sec saturating a 200 MHz system. Your results with the 3Com card are normal for a 16-bit PCMCIA card. The wall you're hitting is the PCMCIA interface. You would have done much better to get a CardBus card (such as the 3CCFE575CT); you should be able to saturate a Fast Ethernet with one of those, and the CPU utilization would be lower. Basically, 10/100 16-bit cards are something of a lie, since you can never get anywhere near 100 Mbps from them. The same is true of 10/100 USB interfaces (the 12 Mbps speed of USB is a hard limit), though USB 2.0 (480 Mbps) could be different if anyone ever produces a USB 2.0 Ethernet interface. > On the Linksys card, I likewise never get more than about 250 KB/sec, > even if the computer is right next to the WAP and the link quality is > 92/92, suggesting that I'm also hitting up against a hard limit. The > WAP is connected to my main system via the motherboard ethernet, which > I know from other means (when I was transferring data from my previous > system to the current one) has no trouble maxing out at 10 MB/sec > (that ethernet does DMA). Not sure about that one; it may be that you're doing about as well as 802.11b and your access point will allow. The bit rate of 802.11b is 11Mbps, but there are other issues with wireless, such as latency (always higher than wired LANs) and delays for the collision avoidance.

April 2, 2002

Re: [Wlug] poor PCMCIA network performance

by Robert L Krawitz

Date: Mon, 1 Apr 2002 10:23:12 -0500 From: "Charles R. Anderson" <cra(a)wpi.edu> Using FTP to check the speed of the network is not a great idea, especially on slow laptop disks. You could be running into the limitations of your disk subsystem. There is a program called ttcp which will test TCP performance with nothing else involved. My laptop disk is fast enough (it's an IBM 32 GB 5400 RPM device); the disk streaming bandwidth (writing to a file) is in excess of 10 MB/sec. The disk on my main system is a SCSI disk that benchmarks at >30 MB/sec. Are those card 16-bit PCMCIA, or 32-bit Cardbus? If they are 16-bit, you are probably maxing out the bus. 16-bit PCMCIA is no better than the old ISA bus in terms of speed. 32-bit Cardbus is like PCI. The wired card is 16 bit. I don't know about the wireless Linksys; I suspect that it is also. For the 10/100 card, check the settings to be sure that both ends of the link are operating at the same duplex. If you are connecting to a repeater (hub, not a switch) you must be using half duplex. If you are connecting to a switch, then the card has to be configured for the same duplex as the switch or you will have horrible performance. Alternatively, the card and the switch can be configured for autonegotiation, so they will both agree on the "best" settings that they both support. Unfortunately, autonegotiation is sometimes buggy and sometimes doesn't work right between different vendor's devices. This is more of a problem with older network devices. I'll look into that. Thanks. -- Robert Krawitz <rlk(a)alum.mit.edu> http://www.tiac.net/users/rlk/ Tall Clubs International -- http://www.tall.org/ or 1-888-IM-TALL-2 Member of the League for Programming Freedom -- mail lpf(a)uunet.uu.net Project lead for Gimp Print/stp -- http://gimp-print.sourceforge.net "Linux doesn't dictate how I work, I dictate how Linux works." --Eric Crampton

April 1, 2002

T&G guotes

by doug waud

Hi I liked the following two quotes in today's (4/1/02) Telegram and Gazette (page A8) "Microsoft bids to acquire Catholic Church" and "Microsoft exploding an underground nuclear device" doug

April 1, 2002

Re: [Wlug] poor PCMCIA network performance

by Charles R. Anderson

Using FTP to check the speed of the network is not a great idea, especially on slow laptop disks. You could be running into the limitations of your disk subsystem. There is a program called ttcp which will test TCP performance with nothing else involved. Are those card 16-bit PCMCIA, or 32-bit Cardbus? If they are 16-bit, you are probably maxing out the bus. 16-bit PCMCIA is no better than the old ISA bus in terms of speed. 32-bit Cardbus is like PCI. For the 10/100 card, check the settings to be sure that both ends of the link are operating at the same duplex. If you are connecting to a repeater (hub, not a switch) you must be using half duplex. If you are connecting to a switch, then the card has to be configured for the same duplex as the switch or you will have horrible performance. Alternatively, the card and the switch can be configured for autonegotiation, so they will both agree on the "best" settings that they both support. Unfortunately, autonegotiation is sometimes buggy and sometimes doesn't work right between different vendor's devices. This is more of a problem with older network devices. On Mon, Apr 01, 2002 at 08:14:06AM -0500, Robert L Krawitz wrote: rlk> I use two network cards in my Dell Inspiron 8000 (PIII-700): a 3Com rlk> 3CCFE574BT and a Linksys wireless card. With the 3Com card (which is rlk> a 10/100) I never get better than 1.4 MB/sec (about 12 Mb/sec), even rlk> when the card is in 100 Mb mode (at some point in the past the limit rlk> was about 2.5 MB/sec). At that rate, top shows my CPU utilization rlk> pegged at about 100%, even doing a simple ftp. I know the card rlk> doesn't do DMA, but that's rather slow even for a PIO transfer. These rlk> cards have been around for ages; I don't think people would have been rlk> too satisfied with 400 KB/sec saturating a 200 MHz system. rlk> rlk> On the Linksys card, I likewise never get more than about 250 KB/sec, rlk> even if the computer is right next to the WAP and the link quality is rlk> 92/92, suggesting that I'm also hitting up against a hard limit. The rlk> WAP is connected to my main system via the motherboard ethernet, which rlk> I know from other means (when I was transferring data from my previous rlk> system to the current one) has no trouble maxing out at 10 MB/sec rlk> (that ethernet does DMA). -- Charles R. Anderson <cra(a)wpi.edu> / http://angus.ind.wpi.edu/~cra/ PGP Key ID: 49BB5886 Fingerprint: EBA3 A106 7C93 FA07 8E15 3AC2 C367 A0F9 49BB 5886

April 1, 2002

RE: [Wlug] T&G guotes

by Keller, Tim

Well, the first one I could actually conceive of... You figure, if they acquire the catholic church they'd just decree that as of baptism you signed the EULA (sorta like opening the package means you agree to the terms (that are on the slip of paper inside the package..)) since the EULA basically means microsoft could set off a nuclear device and get away with it, they'd have no problem proving in court that the church isn't responsible for faulty prietware. As for the nuclear device, If its a microsoft nuclear device I can see the following things happening... 1. Needs repeated reboots before exploding. 2. The trigger will need to have a service pack installed. 3. You'll need to buy MSND 1.5 to get the newest features (like exploding) 4. You'll need to have IE installed along with the latest direct X drivers. 5. before exploding you'll need to get a MSND registration number and enter it in binary using a telegraph pad. 6. "Explode" and "Disarm" will be the same button... Just some thoughts... Tim. -----Original Message----- From: doug waud [mailto:douglas.waud@umassmed.edu] Sent: Monday, April 01, 2002 10:36 AM To: wlug Subject: [Wlug] T&G guotes Hi I liked the following two quotes in today's (4/1/02) Telegram and Gazette (page A8) "Microsoft bids to acquire Catholic Church" and "Microsoft exploding an underground nuclear device" doug _______________________________________________ Wlug mailing list Wlug(a)mail.wlug.org http://mail.wlug.org/mailman/listinfo/wlug

April 1, 2002

poor PCMCIA network performance

by Robert L Krawitz

I use two network cards in my Dell Inspiron 8000 (PIII-700): a 3Com 3CCFE574BT and a Linksys wireless card. With the 3Com card (which is a 10/100) I never get better than 1.4 MB/sec (about 12 Mb/sec), even when the card is in 100 Mb mode (at some point in the past the limit was about 2.5 MB/sec). At that rate, top shows my CPU utilization pegged at about 100%, even doing a simple ftp. I know the card doesn't do DMA, but that's rather slow even for a PIO transfer. These cards have been around for ages; I don't think people would have been too satisfied with 400 KB/sec saturating a 200 MHz system. On the Linksys card, I likewise never get more than about 250 KB/sec, even if the computer is right next to the WAP and the link quality is 92/92, suggesting that I'm also hitting up against a hard limit. The WAP is connected to my main system via the motherboard ethernet, which I know from other means (when I was transferring data from my previous system to the current one) has no trouble maxing out at 10 MB/sec (that ethernet does DMA). eth1 IEEE 802.11-DS ESSID:"..." Nickname:"..." Mode:Managed Frequency:2.437GHz Access Point: 00:04:5A:CF:F3:B1 Bit Rate:11Mb/s Tx-Power=15 dBm Sensitivity:1/3 Retry min limit:8 RTS thr:off Fragment thr:off Power Management:off Link Quality:26/92 Signal level:-76 dBm Noise level:-149 dBm Rx invalid nwid:0 invalid crypt:0 invalid misc:0 I'm running SuSE 7.3 on both systems, with a recompiled kernel based on kernel-source-2.4.16.SuSE-24. I'm not (or don't think I am) running a firewall or NAT on the laptop, despite what's in the module list. I've had no luck finding anything on the net. Modules that look like they should be relevant are: orinoco_cs 4352 1 orinoco 27312 0 [orinoco_cs] hermes 3200 0 [orinoco_cs orinoco] af_packet 11616 0 (autoclean) appletalk 18832 0 (autoclean) ipx 15008 0 (autoclean) ds 6384 2 [orinoco_cs] yenta_socket 8368 2 pcmcia_core 37568 0 [orinoco_cs ds yenta_socket] ipv6 124480 -1 (autoclean) iptable_nat 12560 0 (autoclean) (unused) ip_conntrack 12656 1 (autoclean) [iptable_nat] iptable_filter 1728 0 (autoclean) (unused) ip_tables 10304 4 [iptable_nat iptable_filter] Suggestions? -- Robert Krawitz <rlk(a)alum.mit.edu> http://www.tiac.net/users/rlk/ Tall Clubs International -- http://www.tall.org/ or 1-888-IM-TALL-2 Member of the League for Programming Freedom -- mail lpf(a)uunet.uu.net Project lead for Gimp Print/stp -- http://gimp-print.sourceforge.net "Linux doesn't dictate how I work, I dictate how Linux works." --Eric Crampton

April 1, 2002

Re: [Wlug] Humorous signature file...

by Josh Huber

Chuck Homic <chuck(a)vvisions.com> writes: > Yeah, I think I know at least one person on the list with this > bumper sticker on his car... > > fsck! :) What a dork! ;) -- Josh Huber

March 30, 2002

RE: [Wlug] Request from academic researcher -- follow up

by Clint Moyer

Wow! I cannot even imagine a better example of misdirection... -----Original Message----- From: wlug-admin(a)mail.wlug.org [mailto:wlug-admin@mail.wlug.org]On Behalf Of Utpal M. Dholakia Sent: Friday, March 29, 2002 4:01 PM To: wlug(a)mail.wlug.org Subject: [Wlug] Request from academic researcher -- follow up Dear Linux User Group Moderator: I am a faculty member in the Jones school of management at Rice University in Houston, Texas (this may be verified using a Google search with "Paul Dholakia", or "Utpal Dholakia"). We are conducting an academic research project, studying social interactions online, and how communities organized around products or brands are formed, from a sociological perspective. In this regard, LINUX user groups and mailing lists provide a wonderful context to study these sociological issues. We are conducting a survey of members of LINUX user groups/ mailing lists seeking information on different elements of their thoughts, feelings, and motivations of participating in such groups, and the value of the group to them... The survey is available on the web at the following URL: http://www.surveypro.com/cgi-bin/surveypro/run_survey.cgi?id=2444 Of course, this research is strictly academic, and has no commercial links or interests whatsoever. We hope to publish this research in a sociology journal. As a small token of our appreciation for participants' time, and in the spirit of the philosophy underlying the open-source movement, we will donate $500 total to the five favorite organizations/ groups of our participants' choice (we ask the participant's to nominate their favorite groups in the survey). I am writing to seek your help in conducting this research, first through participating, and second, through encouraging your user group members to participate in the research.. This will not take more than ten minutes of your (and their time). Broad participation will increase the validity and generalizability of our findings, making the research more useful. We welcome all help, and look forward to hearing from you. Best wishes, Paul Dholakia Assistant Professor Rice University http://www.ruf.rice.edu/~dholakia _______________________________________________ Wlug mailing list Wlug(a)mail.wlug.org http://mail.wlug.org/mailman/listinfo/wlug

March 29, 2002

Request from academic researcher -- follow up

by Utpal M. Dholakia

Dear Linux User Group Moderator: I am a faculty member in the Jones school of management at Rice University in Houston, Texas (this may be verified using a Google search with "Paul Dholakia", or "Utpal Dholakia"). We are conducting an academic research project, studying social interactions online, and how communities organized around products or brands are formed, from a sociological perspective. In this regard, LINUX user groups and mailing lists provide a wonderful context to study these sociological issues. We are conducting a survey of members of LINUX user groups/ mailing lists seeking information on different elements of their thoughts, feelings, and motivations of participating in such groups, and the value of the group to them... The survey is available on the web at the following URL: http://www.surveypro.com/cgi-bin/surveypro/run_survey.cgi?id=2444 Of course, this research is strictly academic, and has no commercial links or interests whatsoever. We hope to publish this research in a sociology journal. As a small token of our appreciation for participants' time, and in the spirit of the philosophy underlying the open-source movement, we will donate $500 total to the five favorite organizations/ groups of our participants' choice (we ask the participant's to nominate their favorite groups in the survey). I am writing to seek your help in conducting this research, first through participating, and second, through encouraging your user group members to participate in the research.. This will not take more than ten minutes of your (and their time). Broad participation will increase the validity and generalizability of our findings, making the research more useful. We welcome all help, and look forward to hearing from you. Best wishes, Paul Dholakia Assistant Professor Rice University http://www.ruf.rice.edu/~dholakia

March 29, 2002

Re: [Wlug] Humorous signature file...

by Chuck Homic

On Fri, Mar 29, 2002 at 09:46:01AM -0500, Stephen C. Daukas wrote: > (The first day Microsoft produces something that doesn't suck will be the > day they start making vacuum-cleaners) Yeah, I think I know at least one person on the list with this bumper sticker on his car... fsck! :) -Chuck

March 29, 2002

Humorous signature file...

by Stephen C. Daukas

Came across this one in a newsgroup: (The first day Microsoft produces something that doesn't suck will be the day they start making vacuum-cleaners) Steve

March 29, 2002

Re: [Wlug] Win4Lin

by Wesley Allen

Skip, That's $80 w/o a windows license. Right? Wes Skip Gaede wrote: >Folks, > >This is not a commercial endorsement. I just installed Win4Lin, the product >that lets you run Windows under Linux, and, for those occasions when you need >Windows, it's very nice. Performance is decent, and there are only a few >enhancements needed: > >1) Max memory supported for Windows is 64 MB >2) Only supports serial, parallel, and virtual NIC (no scanners, USB) >3) No DirectX yet > >They're working on adding those goodies. But I can run Photo-Paint, TurboTax, >some Windows-based Bible software packages, ... all without multibooting. > >At $80.00, that's not a bad price to pay. > >--Skip >_______________________________________________ >Wlug mailing list >Wlug(a)mail.wlug.org >http://mail.wlug.org/mailman/listinfo/wlug >

March 28, 2002

Re: [Wlug] Win4Lin

by Skip Gaede

On Thursday 28 March 2002 08:29 am, you wrote: > On 28 Mar 2002, at 7:31, Skip Gaede wrote: > > Folks, > > <snip> > > Granted, I haven't used wine in quite some time, nor do I use the > windows products you mentioned, but doesn't wine support them? > > Though, I remember wine being not fun to setup, so a "plug-and- > play" win emu product is cool > > James > KB1FJQ I looked into buying Photo-Paint for Linux, and as I recall, they had a munged version of wine for their product. Not only that, they had a different munged version of wine to support their latest version of Corel Office (whatever it's called). This version is being supplied to the government, and they have an agreement with Mandrake as well. The web site is http://www.netraverse.com --Skip

March 28, 2002

Re: [Wlug] Win4Lin

by desdinova＠soupforare.net

On 28 Mar 2002, at 7:31, Skip Gaede wrote: > Folks, > <snip> Granted, I haven't used wine in quite some time, nor do I use the windows products you mentioned, but doesn't wine support them? Though, I remember wine being not fun to setup, so a "plug-and- play" win emu product is cool James KB1FJQ > At $80.00, that's not a bad price to pay. > > --Skip > _______________________________________________ > Wlug mailing list > Wlug(a)mail.wlug.org > http://mail.wlug.org/mailman/listinfo/wlug >

March 28, 2002

Win4Lin

by Skip Gaede

Folks, This is not a commercial endorsement. I just installed Win4Lin, the product that lets you run Windows under Linux, and, for those occasions when you need Windows, it's very nice. Performance is decent, and there are only a few enhancements needed: 1) Max memory supported for Windows is 64 MB 2) Only supports serial, parallel, and virtual NIC (no scanners, USB) 3) No DirectX yet They're working on adding those goodies. But I can run Photo-Paint, TurboTax, some Windows-based Bible software packages, ... all without multibooting. At $80.00, that's not a bad price to pay. --Skip

March 28, 2002

Re: [Wlug] SCSI error message

by Skip Gaede

On Monday 25 March 2002 11:59 am, Brian J. Conway wrote: > > Mar 25 09:28:33 pent584 kernel: SCSI subsystem driver Revision: 1.00 > > Mar 25 09:28:33 pent584 kernel: PCI: Found IRQ 10 for device 00:08.0 > > Mar 25 09:28:33 pent584 kernel: (scsi0) <Adaptec AIC-7880 Ultra SCSI host > > adapter> found at PCI 0/8/0 Mar 25 09:28:33 pent584 kernel: (scsi0) Wide > > Channel, SCSI ID=7, 16/255 SCBs Mar 25 09:28:33 pent584 kernel: (scsi0) > > Warning - detected auto-termination Mar 25 09:28:33 pent584 kernel: > > (scsi0) Please verify driver detected settings are correct. Mar 25 > > 09:28:33 pent584 kernel: (scsi0) If not, then please properly set the > > device termination Mar 25 09:28:33 pent584 kernel: (scsi0) in the Adaptec > > SCSI BIOS by hitting CTRL-A when prompted Mar 25 09:28:33 pent584 kernel: > > (scsi0) during machine bootup. > > This just means the card is set to detect if termination is in place. You > can turn it off you have terminators on all the appropriate ends, but it > doesn't really matter. > > > Mar 25 09:28:33 pent584 kernel: (scsi0) Cables present (Int-50 YES, > > Int-68 YES, Ext-68 YES) Mar 25 09:28:33 pent584 kernel: (scsi0) Illegal > > cable configuration!! Only two Mar 25 09:28:33 pent584 kernel: (scsi0) > > connectors on the SCSI controller may be in use at a time! > > Apparently, you can only use 2 of the 3 connectors on the card (external > 68-pin, internal 50-pin, and and internal 68-pin). I've never run into > that one before, myself. You must choose, but choose... wisely. > > -b > > _______________________________________________ > Wlug mailing list > Wlug(a)mail.wlug.org > http://mail.wlug.org/mailman/listinfo/wlug Note: You can attach a narrow drive to a wide cable with a 68 --> 50 pin adapter. I have a wide drive and a narrow tape device and they are both connected to my wide cable...

March 25, 2002

Re: [Wlug] SCSI error message

by Theo Van Dinter

On Mon, Mar 25, 2002 at 11:55:51AM -0500, Theo Van Dinter wrote: > My recollection agrees with this error message, which is basically that > you're not supposed to be using all 3 of the SCSI connectors on the card. To clarify, I used to have a AHA-2940W and found this to be true. I poked at Adaptec's website for you (sorry about the hideously long URL here): http://ask.adaptec.com/cgi-bin/adaptec_tic.cfg/php/enduser/std_adp.php?p_si… Q: I'm thinking of attaching multiple SCSI devices on my SCSI controller. Can I use all connector available on my SCSI controller? Can I use all Internal and External connectors of my SCSI controller simultaneously? This information applies to the following product(s): - AVA-2906, AHA-2930, AHA-2930C, AHA-2930CU, AHA-2940, AHA-2940U, AHA- 2940W, AHA-2944W, AHA-2944UW, AHA-2940UW, AHA-3940 series, AHA-3940U, AHA-3940W, AHA-3940UW,2940U2W, AHA-2930U2, 3950U2, Adaptec SCSI Card- 19160, Adaptec SCSI Card-29160N, Adaptec SCSI Card-29160, Adaptec SCSI Card- 39160 A: It is possible to use all internal and external connectors simultaneously on all currently produced Adaptec's SCSI cards/controllers. Just make sure you have the proper SCSI cables, terminations on your SCSI Bus and devices. Caution: This is with the exception of the AHA-2940W and AHA-2940UW controllers. The only 2940UW series controller to support the use of all three connectors at the same time, was the AHA-2940UW Pro. For informational purposes only, the AHA-2940W and AHA-2940UW legacy SCSI cards/ controllers came with three SCSI connectors (two internal and one external). However You can only use any two connectors simultaneously with these controllers. You can either use one connector internally while using the external connector; or use both internal connectors simultaneously but you must not have any device or cable attached externally. -- Randomly Generated Tagline: A collision is a near-miss.

March 25, 2002

Re: [Wlug] SCSI error message

by Brian J. Conway

> Mar 25 09:28:33 pent584 kernel: SCSI subsystem driver Revision: 1.00 > Mar 25 09:28:33 pent584 kernel: PCI: Found IRQ 10 for device 00:08.0 > Mar 25 09:28:33 pent584 kernel: (scsi0) <Adaptec AIC-7880 Ultra SCSI host adapter> found at PCI 0/8/0 > Mar 25 09:28:33 pent584 kernel: (scsi0) Wide Channel, SCSI ID=7, 16/255 SCBs > Mar 25 09:28:33 pent584 kernel: (scsi0) Warning - detected auto-termination > Mar 25 09:28:33 pent584 kernel: (scsi0) Please verify driver detected settings are correct. > Mar 25 09:28:33 pent584 kernel: (scsi0) If not, then please properly set the device termination > Mar 25 09:28:33 pent584 kernel: (scsi0) in the Adaptec SCSI BIOS by hitting CTRL-A when prompted > Mar 25 09:28:33 pent584 kernel: (scsi0) during machine bootup. This just means the card is set to detect if termination is in place. You can turn it off you have terminators on all the appropriate ends, but it doesn't really matter. > Mar 25 09:28:33 pent584 kernel: (scsi0) Cables present (Int-50 YES, Int-68 YES, Ext-68 YES) > Mar 25 09:28:33 pent584 kernel: (scsi0) Illegal cable configuration!! Only two > Mar 25 09:28:33 pent584 kernel: (scsi0) connectors on the SCSI controller may be in use at a time! Apparently, you can only use 2 of the 3 connectors on the card (external 68-pin, internal 50-pin, and and internal 68-pin). I've never run into that one before, myself. You must choose, but choose... wisely. -b

March 25, 2002

Re: [Wlug] SCSI error message

by Theo Van Dinter

On Mon, Mar 25, 2002 at 11:48:57AM -0500, Bill Mills-Curran wrote: > Mar 25 09:28:33 pent584 kernel: (scsi0) Cables present (Int-50 YES, Int-68 YES, Ext-68 YES) > Mar 25 09:28:33 pent584 kernel: (scsi0) Illegal cable configuration!! Only two > Mar 25 09:28:33 pent584 kernel: (scsi0) connectors on the SCSI controller may be in use at a time! > > > Other than these messages, the disk seems to be working fine. I've > tried different termination settings on the disk, but nothing has > helped. There is a terminator on the scsi cable, and I did not change > it. > > Any ideas? My recollection agrees with this error message, which is basically that you're not supposed to be using all 3 of the SCSI connectors on the card. Are you actually using all three (internal narrow and wide, and the external wide,) or is that an error? If it's an error, then you probably want to check for an auto-termination setting in the SCSI BIOS. -- Randomly Generated Tagline: "How to Boil Water, in 500 easy steps" by Chuck Forsberg.

March 25, 2002

SCSI error message

by Bill Mills-Curran

I added a 2nd disk to my linux box, and I'm getting these messages at boot. (I'm not 100% sure they are new messages, because it's been a long time since I've booted.) Mar 25 09:28:33 pent584 kernel: SCSI subsystem driver Revision: 1.00 Mar 25 09:28:33 pent584 kernel: PCI: Found IRQ 10 for device 00:08.0 Mar 25 09:28:33 pent584 kernel: (scsi0) <Adaptec AIC-7880 Ultra SCSI host adapter> found at PCI 0/8/0 Mar 25 09:28:33 pent584 kernel: (scsi0) Wide Channel, SCSI ID=7, 16/255 SCBs Mar 25 09:28:33 pent584 kernel: (scsi0) Warning - detected auto-termination Mar 25 09:28:33 pent584 kernel: (scsi0) Please verify driver detected settings are correct. Mar 25 09:28:33 pent584 kernel: (scsi0) If not, then please properly set the device termination Mar 25 09:28:33 pent584 kernel: (scsi0) in the Adaptec SCSI BIOS by hitting CTRL-A when prompted Mar 25 09:28:33 pent584 kernel: (scsi0) during machine bootup. Mar 25 09:28:33 pent584 kernel: (scsi0) Cables present (Int-50 YES, Int-68 YES, Ext-68 YES) Mar 25 09:28:33 pent584 kernel: (scsi0) Illegal cable configuration!! Only two Mar 25 09:28:33 pent584 kernel: (scsi0) connectors on the SCSI controller may be in use at a time! Other than these messages, the disk seems to be working fine. I've tried different termination settings on the disk, but nothing has helped. There is a terminator on the scsi cable, and I did not change it. Any ideas? Thanks, Bill

March 25, 2002

Re: [Wlug] normal boot up (was: new mandrake)

by Brad Noyes

Thanks Brian and Jeff for the repsonses. I did remove bootsplash, and no more splash screen. Now when i boot up i don't see any kernel messages or and services start up. I only see message displayed when a service fails. Anyone now how to get my boot to display all the messages? so far the new mandrake 'features' have been somewhat annoying. Thanks, --brad

March 24, 2002

Re: [Wlug] /dev permissions & ownership

by Charles R. Anderson

On Red Hat systems, pam_console is configured by default to change the permissions on certain console devices (floppy, cdrom, sound, etc.) when users login/logout of the console. This is done so the console user automatically can use these devices when they log in, and the permissions revert to root when they log out. This is completely configurable through the /etc/security/console.perms file. On Sat, Mar 23, 2002 at 06:19:05PM -0500, Bill Mills-Curran wrote: subssn594> I just discovered that something has caused many device files (many of subssn594> them audio) have ended up with the wrong ownership/permission. (My subssn594> daughter's account had ownership.) I fixed some, but I have no idea subssn594> what's the right permissions. Is there some way I can check the subssn594> permissions? Ideally, there would be a script that would report which subssn594> device files are not set to the default. (I've changed some, like the subssn594> cdrom and floppy, intentionally.) -- Charles R. Anderson <cra(a)wpi.edu> / http://angus.ind.wpi.edu/~cra/ PGP Key ID: 49BB5886 Fingerprint: EBA3 A106 7C93 FA07 8E15 3AC2 C367 A0F9 49BB 5886

March 24, 2002

Re: [Wlug] /dev permissions & ownership

by Skip Gaede

On Saturday 23 March 2002 06:19 pm, Bill Mills-Curran wrote: > I just discovered that something has caused many device files (many of > them audio) have ended up with the wrong ownership/permission. (My > daughter's account had ownership.) I fixed some, but I have no idea > what's the right permissions. Is there some way I can check the > permissions? Ideally, there would be a script that would report which > device files are not set to the default. (I've changed some, like the > cdrom and floppy, intentionally.) > The script that creates the device files is called MAKEDEV. I don't know what distro you have, but check your CD (or web site) for the script. You may find that it was left in the /dev directory, and all you need to do is execute it. If you use devfs, the devices are created at boot time. --Skip

March 24, 2002

Re: [Wlug] wanted: inexpensive server

by desdinova＠soupforare.net

On 22 Mar 2002, at 17:33, Michelle Vadeboncoeur wrote: <SNIP> > Again, thanks in advance for any pointers you can send! I would suggest an IBM PC Server 704 like her: http://cgi.ebay.com/ws/eBayISAPI.dll?ViewItem&item=2011623032 1/2/4-way PPro, up to three redundant power supplies, 12 hot swappable SCSI bays, and sex appeal, IMHO :) Granted, it's not a whiz-bang 8-way Xeon of doom, but it's a pretty robust platform. James KB1FJQ > -- > --==*==-- --==*==-- Michelle R. Vadeboncoeur --==*==-- > --==*==-- > mrv(a)kluge.net: http://www.kluge.net/~mrv/ > > > _______________________________________________ > Wlug mailing list > Wlug(a)mail.wlug.org > http://mail.wlug.org/mailman/listinfo/wlug >

March 24, 2002

/dev permissions & ownership

by Bill Mills-Curran

I just discovered that something has caused many device files (many of them audio) have ended up with the wrong ownership/permission. (My daughter's account had ownership.) I fixed some, but I have no idea what's the right permissions. Is there some way I can check the permissions? Ideally, there would be a script that would report which device files are not set to the default. (I've changed some, like the cdrom and floppy, intentionally.) TIA, Bill

March 23, 2002