Re: [H] server died again!
At 07:34 PM 27/09/2008, DHSinclair wrote: No. I do NOT know that I have 2 failed drives. That was my posit at this point. Certainly, if I do have 2/3 dead, I would expect to be dead in the water completely. Kiss my last install and all my data goodbye! Am I close? Boo I suspect only one failed drive ATM, but can not yet prove this. Once I can get that far, I do know how to recover. I am not yet there, I believe. And, why I am consulted with the IT Pros of the List. All I have is the RAID alarm, and w2kserver OS that will not complete a boot to its login prompt. It will get to an F8 prompt. I tried an F8 boot but got another BSOD. I may try again and look for a last known. prompt to choose. Barring that, considering the RAID, and my lack of knowledge, I feel I only have the power to do more harm than good. I do have my lan client's now pointed back to the ESET Barn; so future vdefs should continue to arrive auto-magically. (They do. This client is now +2 above the server!) Thank you. Duncan At 18:00 09/27/2008 -0400, you wrote: Wait, you had two drives fail at once? Well, with your setup, you could only survive losing one drive at a time. DHSinclair wrote: Ben, Was running in RAID 5 (?); 2 data drives and a parity drive. With the previous OS install and with the first 2 drive failures, the alarm set and I got a msg at the w2k desktop that the RAID was operating in a degraded mode, but still fully operational. Adaptec SM Pro confirmed which drive was inop. The previous glitches cleared up as soon as I got replacement drives and told SM Pro to rebuild the RAID. This time I can not get to SM Pro. Just my luck. This time, I walked in to a frozen desktop, mouse cursor changed to an UParrow/DWN arrow (never seen this before), a stuck messenger svc window about one of my clients not able to contact the server, and an inop START button - even from kbd commands. I suspect that my 1st mistake was to press the RESET button to reboot. I am not learning server well at all. Thanks for the view of no virus. I will proceed to try and find out which, or, how many of the drives are toast! I have been expecting the oldest of the 3 to fail sometime this year! Thank you. Duncan
Re: [H] server died again!
At 07:34 PM 27/09/2008, DHSinclair wrote: No. I do NOT know that I have 2 failed drives. That was my posit at this point. Certainly, if I do have 2/3 dead, I would expect to be dead in the water completely. Kiss my last install and all my data goodbye! Am I close? I'd boot up with a recovery CD (like Bart's) and then take a look at the drives from there. You may run into RAID compatibility issues, however. T
Re: [H] server died again!
Thane, Yes, I do plan to use my Bart's CD a bit later. I did one check last night b4 bed by booting into the low level bios for the A7499 chip of the RAID card -ctrlA - during the server post. This got me to the low-speed side of the card, but it is here that I can view/scan both of the card's channels- A and B. The B channel holds the card's controller, a CDROM, and my dds tape unit. This is correct and expected. Sadly, when I scanned the A channel (U160) all I saw was the RAID Card's controller. None of the 3 installed and running hds answered the scan. Odd. This I have never seen before in scsi-land. This leads me to posit a bad Channel A cable, the RAID cage (SCA80) interface board, or, the RAID controller card as possible suspects. The hard drives do actually spin up, so I conclude that +12v is working after a fashion. I will check +5v today and carefully inspect the cable, cage and RAID card. No, I have ruled out the server psu either! If nothing shows up, then I will proceed to drive the Bart's CD. This is a weird one for me. Thank you for the shout. Duncan At 10:29 09/28/2008 -0300, you wrote: At 07:34 PM 27/09/2008, DHSinclair wrote: No. I do NOT know that I have 2 failed drives. That was my posit at this point. Certainly, if I do have 2/3 dead, I would expect to be dead in the water completely. Kiss my last install and all my data goodbye! Am I close? I'd boot up with a recovery CD (like Bart's) and then take a look at the drives from there. You may run into RAID compatibility issues, however. T
Re: [H] server died again!
At 12:47 PM 28/09/2008, DHSinclair wrote: Thane, Yes, I do plan to use my Bart's CD a bit later. I did one check last night b4 bed by booting into the low level bios for the A7499 chip of the RAID card -ctrlA - during the server post. This got me to the low-speed side of the card, but it is here that I can view/scan both of the card's channels- A and B. The B channel holds the card's controller, a CDROM, and my dds tape unit. This is correct and expected. Sadly, when I scanned the A channel (U160) all I saw was the RAID Card's controller. None of the 3 installed and running hds answered the scan. Odd. This I have never seen before in scsi-land. This leads me to posit a bad Channel A cable, the RAID cage (SCA80) interface board, or, the RAID controller card as possible suspects. The hard drives do actually spin up, so I conclude that +12v is working after a fashion. I will check +5v today and carefully inspect the cable, cage and RAID card. No, I have ruled out the server psu either! If nothing shows up, then I will proceed to drive the Bart's CD. This is a weird one for me. Thank you for the shout. I'd start by switching SCSI cables, then SCSI connectors. (You've probably already done this, I know.) T
Re: [H] server died again!
Thane, Yes I would suspect the scsi cable(s) first also. No, I have net even opened the server box yet. LOL! Every time I put my hands into this server too soon, I somehow muck something up. Not fear, just some healthy respect... :) Besides this one is so foreign to me in the past 33yr of scsi-land, I wanted to ask the IT gurus their thoughts first. Thank you. More later this afternoon after I dig around a bit. Duncan At 12:57 09/28/2008 -0300, you wrote: At 12:47 PM 28/09/2008, DHSinclair wrote: Thane, Yes, I do plan to use my Bart's CD a bit later. I did one check last night b4 bed by booting into the low level bios for the A7499 chip of the RAID card -ctrlA - during the server post. This got me to the low-speed side of the card, but it is here that I can view/scan both of the card's channels- A and B. The B channel holds the card's controller, a CDROM, and my dds tape unit. This is correct and expected. Sadly, when I scanned the A channel (U160) all I saw was the RAID Card's controller. None of the 3 installed and running hds answered the scan. Odd. This I have never seen before in scsi-land. This leads me to posit a bad Channel A cable, the RAID cage (SCA80) interface board, or, the RAID controller card as possible suspects. The hard drives do actually spin up, so I conclude that +12v is working after a fashion. I will check +5v today and carefully inspect the cable, cage and RAID card. No, I have ruled out the server psu either! If nothing shows up, then I will proceed to drive the Bart's CD. This is a weird one for me. Thank you for the shout. I'd start by switching SCSI cables, then SCSI connectors. (You've probably already done this, I know.) T
[H] server died again #2
The server troubleshooting is on. I took the raid-cage loose from the server and found big time dust bunnies. Hmm. Seems I missed my June PM-another machine crash as my records show. Power on-goes to previously reported BSOD. OK, not dust bunnies. I then replaced the main scsi cable from the Adaptec controller to the raid-cage with a same-same cable removed from another working scsi client. Power on-goes to previously reported BSOD. Hmm.perhaps not the cable? I do not have a spare Adaptec 3200S. Not even sure if it is still available. Still confused by why Windows will not complete a boot w/raid in a degraded mode. So, I decide to try and boot using all combinations of 2 of the 3 drives. This appears to now point to a potential failed drive (0:5). This is my new failure matrix: Make drive 0:1 and 0:3 hot: server completes post, windows starts boot, displays the bottom screen ruler, W2K opening screen appears and the progress meter completes, screen then goes to the previously reported BSOD. Make drive 0:3 and 0:5 hot: server completes post, black screen w/ -Operating System not found. Make drive 0:1 and 0:5 hot: server completes post, black screen w/ -Operating System not found. I have even checked this logic letting the drives use my other two spare slots of 2 and 4. The results are the same. So I do think at least my raid cage is OK. Yes, the Adaptec controller could still be suspect. ATM, my conclusion is that the drive in position 0:5 is toast. Even though it spins up and appears to answer commands (yellow LED) along with the other two raid members. Drive 0:5's yellow LED only stops blinking once the Adaptec controller sets the alarm on. The windows boot continues to the eventual BSOD. I remain confused about why Windows will not complete a boot with a degraded raid as I've been told it should. Any ideas would be helpful. I am sure I followed all the proper steps when I rebuilt this server some time back. In the am I will begin the drive replacement business. Thank you for any suggestions. Duncan
Re: [H] White list or black list?
On Sat, 27 Sep 2008, mark.dodge wrote: I need to limit which web pages multiple business computers can go to. What is the file that does this? I want to limit IE to around 12 sites the business needs for insurance and their page and a couple of medical research sites. I've done a google on whitelist but it comes up with programs that can or can't be run. Multiple methods, if you can fit all of the websites into the Don't use a proxy for these sites line in the internet explorer proxy settings you can setup a dummy proxy and use that line to whitelist your sites. I wouldn't just do it at the client, I would add it as a group policy and make it so they can't change the proxy server. Probably the best way of doing this is removing their default gateway and setting up an actual content filtering proxy server. The Cyclope Filter I recommended previously works well, but costs per user. You can quickly and easilly add and remove sites from the whitelist and blacklist in the proxy, and it can be configured in either mode (Whitelist or Blacklist). It doesn't have individual user settings, so you'd need something a bit more powerful if you're looking for allowing different users to access different sites. It all depends on how determined your users are to get past the filter. If you allow access out to the internet on other ports, they can always install an SSH server somewhere, use an SSH client and tunnel all their HTTP requests to a proxy server they setup elsewhere, and bypass. That's what removing their default gateway fixes. Christopher Fisk -- Leela: Hey, you know what might be a hoot? Professor: No. Why would I know that? -- This message has been scanned for viruses and dangerous content by MailScanner, and is believed to be clean.
Re: [H] server died again #2
Windows, or any other OS for that matter, will boot successfully from a degraded array. Windows itself has no knowledge of the array being optimal, degraded, or anything else--that's handled by the controller itself. The operating system doesn't even know anything about the individual disks attached, since again, it's all obfuscated by the controller's firmware. My guess is that your bad drive has been going bad for a while, and has caused some silent data corruption prior to fully kicking the bucket. Or, the card itself is bad. Or, the driver is causing some problems as it deals with/attempts to report the fault condition. RAID is never a substitute for good backup procedures. RAID does nothing more than increase availability (and performance, depending upon a number of factors). It's no data-protection cure-all, as your situation may ultimately reveal. Greg I remain confused about why Windows will not complete a boot with a degraded raid as I've been told it should. Any ideas would be helpful. I am sure I followed all the proper steps when I rebuilt this server some time back. In the am I will begin the drive replacement business. Thank you for any suggestions. Duncan