RE: [leaf-user] Compact Flash Boot Failure
Hi! IMHO using dd to backup and restore CF's is a very very bad idea. Specially if these CF are from different manufacturers, often with different CHS settings. I never had a problem with CF, as I syslinux them and copy the LRP's by hand. The last 3 or 4 updates I did on my router were even done with a CF USB adapter from Windoze... Hope this helps a bit. Luis Correia Bering uClibc Team Member PGP Fingerprint: BC44 D7DA 5A17 F92A CA21 9ABE DFF0 3540 2322 21F6 Key Server: http://pgp.mit.edu -Original Message- From: Brock Nanson [mailto:[EMAIL PROTECTED] Sent: Saturday, January 08, 2005 1:01 AM To: Erich Titl Cc: leaf-user-lists.sourceforge.net Subject: Re: [leaf-user] Compact Flash Boot Failure Hi Erich, I've sent this box back to the user. Not having all the hardware available that I did when I first built the boxes was a problem... specifically, no Linux laptop with PCMCIA. Finally pulled a Knoppix ISO and got an XP laptop going off the CD-ROM drive (note to anyone that cares, an Inspiron 1150 doesn't like the default Knoppix boot. Specifying the 2.6 kernel at boot solves that)! What I did: copied all the packages off the CF for safe keeping (had a copy from some months ago, but wasn't sure what changes might have been made along the way). syslinux -s /dev/hde1 Put it back in the box. Booted to syslinux (with the BIOS set manually to what I *thought* was the closest settings) but not beyond... loading linux, loading init, then nothing. Copied the packages and config file back onto the CF, overwriting what was there. Back in the router. Boot was fine this time. During the startup screens I see the CF detected by Linux as CHS 244/4/32 (16MB Lexar). I was set manually to 251/2/32 which works for the 8MB version of this card. The image on the 16 was dd'd from an 8MB image, so this might (?) be why it worked... If I'm honest, I'll 'fess up and admit I forgot to reset the CHS after messing with an 8MB card! ;-) Manually changed the BIOS to 244/4/32 and rebooted several times. No issues that I can see. Now, when the BIOS autodetected the card originally, it came up with Pre =-1, LZ=244. These parameters are blank when manually configured. Do they make any difference to the CF? Oh yeah, I also put the whole mess into a duplicate machine (same model) to test the power supply theory. New CF adapter too. Thanks to all who made suggestions! Brock Erich Titl wrote: Brock Brock Nanson wrote: Hi Erich, Well, there isn't much I can tell you. The box in question resides 8 hours away. It worked for approximately 6 months, then I got the call from that office to tell me it wouldn't boot. They sent the CF back to me and I poked around it a little. Not much to see, it simply wouldn't boot. So I assumed a bad CF, rebuilt their system on a new card and sent it back to them. They plugged the new card into the box and it has worked properly since mid-May, failing just before Christmas. Same symptoms. This time I had them send the whole box back to me. I don't see anything wrong with the box. However, as this happens so rarely, it could really be anything. Flakey power supply, two bad CF's in a row (possible if not very likely) I had symptoms like that on BiWin CF's. I have a number of those deployed and I usually check them by writing the entire CF. If they report a problem back they go... It might be the CF adapter, yu did not change that one did you? IO at boot is conrtolled by BIOS, not a linux driver, so this may be a difference too. , maybe even an unwelcome visitor... who knows. Possible, or else having the CF mounted at a full back up? you could write an md5sum of the contents and check regularly. When it takes 6 or 7 months to repeat the failure, it's difficult to efficiently trouble shoot this problem. ;-) Unfortunately, the symptoms aren't much to go on! It worked one day, but not the next! I'm going to use a twin machine to replace this one, with a new (same model PC-Engines adapter board - type 1C) CF adapter and CF card. If this one fails later this year I'll be back where I started... pretty much nowhere. I just don't understand what could wipe the boot area twice in a row, but leave the rest of the card untouched... power spikes, flakey IDE cable, nothing really adds up to this result. My only idea revolves around the possibility that the BIOS is not auto-detecting the card properly. I've talked to Lexar, OK so you are using Lexar, better than BiWin IMHO. Good Luck Erich --- The SF.Net email is sponsored by: Beat the post-holiday blues Get a FREE limited edition SourceForge.net t-shirt from ThinkGeek. It's fun and FREE
Re: [leaf-user] Compact Flash Boot Failure
Brock Brock Nanson wrote: Hi Erich, Well, there isn't much I can tell you. The box in question resides 8 hours away. It worked for approximately 6 months, then I got the call from that office to tell me it wouldn't boot. They sent the CF back to me and I poked around it a little. Not much to see, it simply wouldn't boot. So I assumed a bad CF, rebuilt their system on a new card and sent it back to them. They plugged the new card into the box and it has worked properly since mid-May, failing just before Christmas. Same symptoms. This time I had them send the whole box back to me. I don't see anything wrong with the box. However, as this happens so rarely, it could really be anything. Flakey power supply, two bad CF's in a row (possible if not very likely) I had symptoms like that on BiWin CF's. I have a number of those deployed and I usually check them by writing the entire CF. If they report a problem back they go... It might be the CF adapter, yu did not change that one did you? IO at boot is conrtolled by BIOS, not a linux driver, so this may be a difference too. , maybe even an unwelcome visitor... who knows. Possible, or else having the CF mounted at a full back up? you could write an md5sum of the contents and check regularly. When it takes 6 or 7 months to repeat the failure, it's difficult to efficiently trouble shoot this problem. ;-) Unfortunately, the symptoms aren't much to go on! It worked one day, but not the next! I'm going to use a twin machine to replace this one, with a new (same model PC-Engines adapter board - type 1C) CF adapter and CF card. If this one fails later this year I'll be back where I started... pretty much nowhere. I just don't understand what could wipe the boot area twice in a row, but leave the rest of the card untouched... power spikes, flakey IDE cable, nothing really adds up to this result. My only idea revolves around the possibility that the BIOS is not auto-detecting the card properly. I've talked to Lexar, OK so you are using Lexar, better than BiWin IMHO. Good Luck Erich --- The SF.Net email is sponsored by: Beat the post-holiday blues Get a FREE limited edition SourceForge.net t-shirt from ThinkGeek. It's fun and FREE -- well, almosthttp://www.thinkgeek.com/sfshirt leaf-user mailing list: leaf-user@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/leaf-user SR FAQ: http://leaf-project.org/pub/doc/docmanager/docid_1891.html
Re: [leaf-user] Compact Flash Boot Failure
Hi, The PCMCIA flash technology was something I used before CF was available. I have some PCMCIA to IDE adapters that I bought from Addtron, and I bought a bunch of 32M PCMCIA cards. The solution was fairly expensive at about $200 per system (compare that to a $20 adapter and a $10 CF card today), but at the time it was more about the reliability in my router than anything else. The only reason that I mentioned it was that I have systems that have been in place for 5 years and have never had a problem. I've had some of them running since when Dave and the gang were starting out with LRP. Thanks, Jim wing newton wrote: Jim, How did you get it booted from pcmcia flash ? Thanks. Newton --- Jim Walters [EMAIL PROTECTED] wrote: Hi, I am running a number of boxes for years from CF and PCMCIA flash and have never had anything like that happen. Without any more information, I would tend to look at the power supply as a potential causer of problems, since bad or slow powerup during the startup phase could confuse the CF cards. I realize that is significantly less than $0.02 worth, but it's all I've got. Thanks, Jim Walters Brock Nanson wrote: I've got several boxes running several versions of Bering and Bering uClibc, all on Compact Flash cards. They've all been reliable, except for one box. Twice in the last six months, the box has failed to boot. The CF is recognized in the BIOS, but no boot information seems to be found. The first time it happened, I swapped the CF card for another and built the system from scratch. This time I'm wondering if the computer could be part of the problem. I don't like the double coincidences of it a) failing in the first place and b) failing *twice* in the same box on different CF's. Especially when all the others have run for years. Has anyone had a problem like this? I'd like to know if the box could be the problem before I take the time to replace this CF yet again. The unit is in a remote location and is fairly mission-critical. I can't have this happen so regularly. Brock --- The SF.Net email is sponsored by: Beat the post-holiday blues Get a FREE limited edition SourceForge.net t-shirt from ThinkGeek. It's fun and FREE -- well, almosthttp://www.thinkgeek.com/sfshirt leaf-user mailing list: leaf-user@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/leaf-user SR FAQ: http://leaf-project.org/pub/doc/docmanager/docid_1891.html --- The SF.Net email is sponsored by: Beat the post-holiday blues Get a FREE limited edition SourceForge.net t-shirt from ThinkGeek. It's fun and FREE -- well, almosthttp://www.thinkgeek.com/sfshirt leaf-user mailing list: leaf-user@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/leaf-user SR FAQ: http://leaf-project.org/pub/doc/docmanager/docid_1891.html __ Do you Yahoo!? All your favorites on one personal page Try My Yahoo! http://my.yahoo.com --- The SF.Net email is sponsored by: Beat the post-holiday blues Get a FREE limited edition SourceForge.net t-shirt from ThinkGeek. It's fun and FREE -- well, almosthttp://www.thinkgeek.com/sfshirt leaf-user mailing list: leaf-user@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/leaf-user SR FAQ: http://leaf-project.org/pub/doc/docmanager/docid_1891.html
Re: [leaf-user] Compact Flash Boot Failure
Brock Brock Nanson wrote: I've got several boxes running several versions of Bering and Bering uClibc, all on Compact Flash cards. They've all been reliable, except for one box. Twice in the last six months, the box has failed to boot. The CF is recognized in the BIOS, but no boot information seems to be found. The first time it happened, I swapped the CF card for another and built the system from scratch. This time I'm wondering if the computer could be part of the problem. I don't like the double coincidences of it a) failing in the first place and b) failing *twice* in the same box on different CF's. Especially when all the others have run for years. Has anyone had a problem like this? I'd like to know if the box could be the problem before I take the time to replace this CF yet again. The unit is in a remote location and is fairly mission-critical. I can't have this happen so regularly. Yes, I have seen this before, mostly caused by bad CF's, sometimes by bad saves of the entire system.Would you mind to give us details about the failure Erich --- The SF.Net email is sponsored by: Beat the post-holiday blues Get a FREE limited edition SourceForge.net t-shirt from ThinkGeek. It's fun and FREE -- well, almosthttp://www.thinkgeek.com/sfshirt leaf-user mailing list: leaf-user@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/leaf-user SR FAQ: http://leaf-project.org/pub/doc/docmanager/docid_1891.html
Re: [leaf-user] Compact Flash Boot Failure
Hi Erich, Well, there isn't much I can tell you. The box in question resides 8 hours away. It worked for approximately 6 months, then I got the call from that office to tell me it wouldn't boot. They sent the CF back to me and I poked around it a little. Not much to see, it simply wouldn't boot. So I assumed a bad CF, rebuilt their system on a new card and sent it back to them. They plugged the new card into the box and it has worked properly since mid-May, failing just before Christmas. Same symptoms. This time I had them send the whole box back to me. I don't see anything wrong with the box. However, as this happens so rarely, it could really be anything. Flakey power supply, two bad CF's in a row (possible if not very likely), maybe even an unwelcome visitor... who knows. When it takes 6 or 7 months to repeat the failure, it's difficult to efficiently trouble shoot this problem. ;-) Unfortunately, the symptoms aren't much to go on! It worked one day, but not the next! I'm going to use a twin machine to replace this one, with a new (same model PC-Engines adapter board - type 1C) CF adapter and CF card. If this one fails later this year I'll be back where I started... pretty much nowhere. I just don't understand what could wipe the boot area twice in a row, but leave the rest of the card untouched... power spikes, flakey IDE cable, nothing really adds up to this result. My only idea revolves around the possibility that the BIOS is not auto-detecting the card properly. I've talked to Lexar, but so far they have not produced any CHS information for their cards. And I can't find anything useful on the 'net. I'm thinking I should manually configure the CHS settings, just in case it is being auto-detected incorrectly. Brock Erich Titl wrote: Brock Brock Nanson wrote: I've got several boxes running several versions of Bering and Bering uClibc, all on Compact Flash cards. They've all been reliable, except for one box. Twice in the last six months, the box has failed to boot. The CF is recognized in the BIOS, but no boot information seems to be found. The first time it happened, I swapped the CF card for another and built the system from scratch. This time I'm wondering if the computer could be part of the problem. I don't like the double coincidences of it a) failing in the first place and b) failing *twice* in the same box on different CF's. Especially when all the others have run for years. Has anyone had a problem like this? I'd like to know if the box could be the problem before I take the time to replace this CF yet again. The unit is in a remote location and is fairly mission-critical. I can't have this happen so regularly. Yes, I have seen this before, mostly caused by bad CF's, sometimes by bad saves of the entire system.Would you mind to give us details about the failure Erich --- The SF.Net email is sponsored by: Beat the post-holiday blues Get a FREE limited edition SourceForge.net t-shirt from ThinkGeek. It's fun and FREE -- well, almosthttp://www.thinkgeek.com/sfshirt leaf-user mailing list: leaf-user@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/leaf-user SR FAQ: http://leaf-project.org/pub/doc/docmanager/docid_1891.html
Re: [leaf-user] Compact Flash Boot Failure
Hi, I am running a number of boxes for years from CF and PCMCIA flash and have never had anything like that happen. Without any more information, I would tend to look at the power supply as a potential causer of problems, since bad or slow powerup during the startup phase could confuse the CF cards. I realize that is significantly less than $0.02 worth, but it's all I've got. Thanks, Jim Walters Brock Nanson wrote: I've got several boxes running several versions of Bering and Bering uClibc, all on Compact Flash cards. They've all been reliable, except for one box. Twice in the last six months, the box has failed to boot. The CF is recognized in the BIOS, but no boot information seems to be found. The first time it happened, I swapped the CF card for another and built the system from scratch. This time I'm wondering if the computer could be part of the problem. I don't like the double coincidences of it a) failing in the first place and b) failing *twice* in the same box on different CF's. Especially when all the others have run for years. Has anyone had a problem like this? I'd like to know if the box could be the problem before I take the time to replace this CF yet again. The unit is in a remote location and is fairly mission-critical. I can't have this happen so regularly. Brock --- The SF.Net email is sponsored by: Beat the post-holiday blues Get a FREE limited edition SourceForge.net t-shirt from ThinkGeek. It's fun and FREE -- well, almosthttp://www.thinkgeek.com/sfshirt leaf-user mailing list: leaf-user@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/leaf-user SR FAQ: http://leaf-project.org/pub/doc/docmanager/docid_1891.html --- The SF.Net email is sponsored by: Beat the post-holiday blues Get a FREE limited edition SourceForge.net t-shirt from ThinkGeek. It's fun and FREE -- well, almosthttp://www.thinkgeek.com/sfshirt leaf-user mailing list: leaf-user@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/leaf-user SR FAQ: http://leaf-project.org/pub/doc/docmanager/docid_1891.html
RE: [leaf-user] Compact Flash Boot Failure
Hi Brock, Has anyone had a problem like this? I'd like to know if the box could be the problem before I take the time to replace this CF yet again. The unit is in a remote location and is fairly mission-critical. I can't have this happen so regularly. This has happened to me. I am using LEAF Bering-uClibC with bgpd for 4 solid-state routers. Unfortunately one of my routers seems to regularly chew up its CF card. I've switched the types of cards around in the system to no effect. My solution so far has been to back up the drive regularly and not reboot, not exactly a nice solution. I have not had the time to look into this matter properly, but a few ideas that have come to mind are: - Measure the voltage being given to the CF/IDE card. - Measure the voltage the CF/IDE card is giving the CF. - Check the cylinder count etc. on your IDE card, if possible. Rumor has it incorrect IDE setup can lead to this problem. - Replace the CF/IDE and CF card. I have used netcat and disk dump to good effect in backing up / restoring my images. Try: (on destination) : nc -l -p port | dd of=/dev/hda (on original server) : dd if=/dev/hda | nc addr of other machine port as above Regards, P --- The SF.Net email is sponsored by: Beat the post-holiday blues Get a FREE limited edition SourceForge.net t-shirt from ThinkGeek. It's fun and FREE -- well, almosthttp://www.thinkgeek.com/sfshirt leaf-user mailing list: leaf-user@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/leaf-user SR FAQ: http://leaf-project.org/pub/doc/docmanager/docid_1891.html
