Emprex 1gb flash driveI bought a 1gb flash disk from Emprex (becaue it was on sale). I copied some files to it. Some of the files got corrupted. I tried to reformat the drive. The reformat failed. I tried under linux as well. The reformat failed. I am unable to reformat it, and I am unable to copy files to it. I sent their tech support email. They said return it to the store I got it. I told them it was Fry's and they only accept returns for one month. I called their tech support department. They sent me a RMA form to fill out, and they asked for a copy of the sales receipt. I got a replacement drive, and it works normally. Sandisk 2gb flash driveI recently bought a Sandisk flash drive, because it was on sale, and they have a well known name. It came with U3 software which is supposed to allow data on the flash drive to be protected. I tried copying 5 files that were approximately 350mbytes onto the drive. It was very slow. After a while, I got an error message about a failed write. I thought it might have something to do with the U3 software, so I tried to remove it. It cleverly reinstalled itself. I downloaded the Sandisk program to remove the U3 software. It told me that more than one drive was plugged into the USB on my computer. I unplugged everything, and got the same message. I tried the program on my notebook, and it worked. It turned out that I had an internal flash reader plugged into my USB on my main computer. It seems a stupid limitation that the program can't find or allow you to specify where the Sandisk flash drive is. So, I tried copying the files again. It failed the same way. I called Sandisk tech support. They said all kinds of irrelavent things. They asked if I was copying it to the root directory, and said there might be a problem with that. They asked if I had any cdrom burning software and said there might a problem with that. They asked how big the files were and said there might be a problem with that. They had me reformat the flash drive. Finally, I said this was all bull, that FAT32 allows for pretty much any size file any where on the filesystem. I grabbed another 2gb flash drive which I happened to have lying around, and successfully copies the files on to that. Finally they admitted that it might be possible that their brand new flash drive might be defective. They said I could RMA it, but it would be faster to return it to where I bought it. I agreed. I know that inexpensive hardware can fail, but having it not work from the start, where a simple test would reveal the problem is inexcusable. I am very disappointed with the quality control of SanDisk. I am also very disappointed with the quality of their tech support. The only thing they didn't claim could cause a problem was the direction I was pointing the flash drive. They should not mention bogus things. It can easily confuse the less technically adept. |
Three months ago, I bought a PCI 4mb trident chipset video card from Fry's, because it had a $20 rebate. The rebate process was unique in my experience. You had to send a self-addressed stamped envelope. You had to register on their web page first. You had to include a hardcopy of their web page registration. Perhaps this was all designed to minimize the number of rebates submitted.
In any event, I got the money about 14 weeks later. The week after I tried to install the video card. It was dead on arrival. No problem, I thought, the card has a one year warranty. I went to the webpage and filled out their form. I got a reply back from their mailer-daemon, saying that there was no such user as services@kasercorp.com. Clearly their web page mailed the form to some incorrect address. No problem, I thought. I mailed it to support@kasercorp.com, and got back a reply saying "you need to return from you purchase from. We won't take any service to indirect customer besides the tech support." Other than the bad english, I took it to mean. they don't deal with it; its Fry's problem.
I sent them back email, saying Fry's won't deal with it after a month. They never replied to my email again. I called them up, and they repeated their 'take it back to where you got it' line. I called Fry's and they confirmed I had to send it back to the manufacturer. I called Kasercorp again, saying I had spoken to Fry's, and the repeated their line. I called Fry's again, and told them to contact Kasercorp, and gave them their number, and suggested they should not sell products when the manufacturer doesn't stand behind their product and wants Fry's to deal with it.
I called Kasercorp again, and they finally agreed to take the video card back, but I would have to pay shipping. I told them that I fully expected to pay the shipping to return it. No, they said, you have to pay Kasercorp's shipping and handling. I asked them how much it was, and they said it was $10. I told them that it would cost me $5 to ship it to them, and that I could get a new card for $15, so there one year warranty was not much of a warranty at all.
They told me they charged $10 for all their shipping & handling, no matter how much the product cost. I find it really odd that they changed their policy after I called them three times. I explained that I have returned stuff to Plextor, Toshiba, Sony, Adaptec, and others and none of them have charged me shipping and handling from their end. Kasercorp said that is how they do business.
So, I am dismayed with Kasercorp. Their rebate process is a bigger hassle than any other company I have dealt with. They shipped a board with no quality control, as it was dead on arrival. They had a web page for warranty service that doesn't work. They ignored my second email, when I asked them how to get the card serviced. I had to call them three timed before they agreed to service their defective card, and they have no toll-free number. They finally agreed to fix or replace my card, but wanted to charge me probably more than they sell a new card for wholesale. The good news is that pricewatch.com doesn't list any products sold by Kaser. I suspect they are a really low end company that only sells to Fry's. Therefore, I don't recommend buying any of their products under any conditions, including free after rebate, because of all of the above problems.
I bought a Gateway P-120 computer in 1995. Around 2000 it became unreliable. I had to power it several times before it would boot. A little while later, it blew up literally. An epoxy cased IC smoked, and left a crater showing the die.
I bought two Gateway 6400 Server computers in June 2002. Other than the defective design of the Asus CUR-DLS motherboard using the Serverworks LE chipset as detailed in my link, the motherboards were fine. Until one of them broke in 2004. The motherboard would recognize one of the two processors, but not both. As most operating systems are either SMP or non-SMP, the SMP operating systems would not boot. I was able to boot a single processor copy of Linux I had on a hard drive. The board was replaced under warranty.
Plextor has a reputation for making the best cd and dvd burners. I bought a PX-716A burner. It failed after about 3 years of use. Plextor doesn't repair them. The new Plextor drives don't work with Plextools, which can tell you about the C1/C2, and P1/P0 errors of a burned cd or dvd.
I got this Hi-Val drive, because an old x16 drive broke (and there was a rebate on the Hi-Val). Well after 5 weeks of light use the Hi-Val drive broke. I got another one, and the first day it started making noises that sounded like something scraping against the disk itself. Like something from the shower scene from Psycho. I returned it and got the Kenwood True 42x instead.
I have a Toshiba SCSI CDROM drive. It completely failed once. I sent it back to Toshiba, and they fixed it. Sometimes when the machine is booting, the 'busy' light comes on, and stays on when a disc is inside, but that doesn't happen often. People say that Toshiba and Plextor make the best SCSI drives...
Click on the link to read about a really inferior product with really inferior service. Netgear RT-314 Internet Router
Recently, I purchased a Belkin router. I have a bunch of other Belkin stuff, and it all works fine. This router does not have a print server, but I already have an Ethernet connection on my big printer. Included in the case is a 4 port switch. This new router has stateful packet inspection and NAT, which makes for a pretty robust firewall.
The plastic case looks unremarkable. It was easy to configure, and seemed to work well. After a few days of use, the DNS server seemed to fail. From a DOS box, I was unable to successfully ping www.ibm.com (my standard host to ping). I did the kludgey windows "ipconfig/release and then ipconfig/renew" trick, and it didn't help. I rebooted my machine, and everything was working fine.
Roughly every 24 hours, I have the same problem. Sometimes, the release/renew thing fixed the problem. Sometimes not. I sent email to Belkin tech support. They said download a new version of the router software. Unfortunately, it was for a different Belkin router. Fortunately it was smart enough to not install. After a few more go-rounds with Belkin tech support, they said to call them on the phone.
The guy I spoke to was pretty unhelpful. He said to manually set my network card speed. I told him it worked just fine with another router. He wanted to set it anyway. I told him I didn't want to do so. He said all he could suggest was to set a fixed IP address for my machine. I told him I had 10 machines, and I didn't want to have to set the IP addresses for all of them. He said I could set the DNS address manually, and I told him that the purpose for the DHCP server inside the router was to not have to do this. I asked him if there was a new version of the software for the router, and he said no, there were no problems with the existing software. I told him there were problems, because I was having them. He was pretty unhelpful.
It turned out that the problem was not with the Belkin router, but with ZoneAlarm firewall software. See ZoneAlarm for more details. I was unimpressed with Belkin tech support, but the Belkin hardware is pretty reliable.
In Sep 2005, my Belkin router became unreliable. I had to reset it every day or so. I called tech support, and they had good ideas. They had me reset the router, but the problem persisted. I had switched to my DLink Router and it worked fine for over a week. I called Belkin back and told them. We tried the Belkin router again. Same problem. Switched back to DLink, and it worked. Belkin sent me a new router. It has the same part number, as the old router, but it about half the size. Didn't fit in my network closet as it was so small. I need to drill new holes to secure it. Belkin didn't ask for the old one back. It has worked with no issues. First rate technical support. This demonstrates a good reason to have a spare router.
I bought a Best Data CMX110 cable modem around 2001. In Dec 2005, I started having problems downloading data from the Internet. Thanks to some help from Comcast, they were able to determine that 20% of the pings were failing and that the ping times were very long. They suggested my cable modem might be suspect. I thought it was a silly idea, as it was hard to imagine that it could partially fail. I bought a motorola surfboard cable modem. It solved all of my problems. The Best Data modem had partially failed. I have not investigated the cause of the failure yet. Since I have to call Comcast to enable a new cable modem, it is not trivial to debug.
It worked fine for about a year. Then it failed. I called the cable company and after lots of fussing around it started working again. Failed completely one week later. Replaced under warranty. I called around 9pm and someone quickly answered and as soon as I told them it was a retail unit, they gave me return instructions. No EVGA bull*hit about needing the sales receipt, or having to register online. I did have to pay return shipping, but that is pretty standard for retail goods.
I bought two Gateway 6400 Server computers in June 2002. Servers are supposed to be built of quality parts, and have few failures. One of the computers was having a hard time booting. The fans would spin up, but the computer would not boot (power on self test). Sometimes it would boot successfully. After a few months, it failed to boot. I suspected the power supply (Astek model SA320-3525) might be had. I put in a different power supply, and have not had a problem booting.
I bought this when I didn't know much about power supplies. Worked for a few months, and then started failing intermittently. Eventually failed completely.
I bought a 900VA UPS (really 500W which is pretty small for 900VA) from Conext. Conext is the less expensive brand name for APC, which is a very well regarded UPS company. Although I bought in around Nov-20-2004, it was made around Mar-2001. (UPS's are typically shipped with the battery disconnected, and the user connects the battery). It seemed fine. Then the power failed. There was a really horrible noise coming from my computer room. I unplugged everything in the room. I removed all the UPS's from the room, one at a time. It turned out the source of the noise was the Conext UPS. No doubt it was a low battery alarm or something. There was nothing I could to silence the UPS. I suppose I could have removed the battery, but instead I moved the UPS to the garage. It would be nice if there was some way to turn off the alarm. Perhaps the inability to turn off the alarm is related to the complete failure of the UPS. It seemed dead. Nothing I could do would make it power up. I called the support line, and got a very smart tech right away. He had me reset it, to no avail. I mentioned the software was old and didn't support Windows XP. Their web site said to use the default Windows UPS software. The tech explained how to use the APC Power Chute Business software with the UPS. He agreed to cross ship me a replacement UPS as mine was less than two months old. He speculated that the UPS failed because the battery was old. It would be nice if it simply notified the user that the battery was old, instead of turning off completely, but I suppose it would get the users attention...
I acquired a HP scanner. It worked fine. After a few years it failed. I didn't know why. After a little effort, I found the 12v power supply, often known as a wall-wart, had failed. It had a standard 5mm coaxial plug, but it was a non-standard depth. Fortunately, I had a spare power supply from my failed Best Data Cable Modem. A ugly splice job later, and the cord from the HP wall-wart was attached to the cable of the Cable Modem. Fixed the problem, the scanner is now working fine.
For my job, I purchased 23 Dell Linux boxes. They all had the same Elsa video cards. After a few months, one of the displays started looking funny. There were dark horizontal stripes across the screen, wherever there were dark areas on the screen (like the tops of Windows). When you moved the Windows up or down, the dark horizontal stripes moved up and down in sync.
The first Dell support I talked to was pretty clueless. He had me swap monitors, and wanted me to run some Windows diagnostics. I finally convinced him it was the video card. They shipped me a new one and the problem was solved. But then another video card failed the same way. And then another and another. I have lost count, but in about a year, at least 25 cards have failed. Some of the replacement cards have also failed. All in the same way.
Several times I asked Dell to replace all the cards, but they said that their contract includes replacing bad hardware. I pointed out it was much easier to have planned downtime for our Linux boxes, and they repeated that their contract dealt with bad hardware, and until the boards failed, they weren't bad.
We have also had one motherboard and one hard drive fail. Not so bad for equipment that is on 24*7.
I bought several EVGA video cards because they were inexpensive and they used passive heat sinks so they don't make noise. I was using one of the cards on my computer I have hooked up to my TV to watch video. Well, after a few months, the color started failing on the screen. It was intermittent for a few days, but then it failed completely. The strange thing, was during booting there was color, but at higher resolution, there was no color. Clearly the S-Video cable was ok.
I replaced the card with another EVGA card, and everything worked again. I contacted EVGA to get the broken card fixed. After a few days, they replied telling me that I had to register the card on their website before I could get a RMA. I replied that their registration was broken, and I gave them the specific error message that it was producing. I was told that they knew about that problem, and they were working on it, and that I would have to register the card before I could get a RMA. After a few days, the web registration was working, and I registered the card and I got a RMA for it. I sent the card in. It hasn't come back after two weeks. I presume it will eventually come back. I have never had to register in order to get defective equipment repaired. I also had to provide a copy of the sales receipt. Fortunately I keep such things. Adaptec didn't ask for a sales receipt. Neither did Garmin, or any other computer company I have dealt with. I am unimpressed with the customer service of EVGA, and I question the quality of their product.
I have had enough hard drives that I can make some sense about their reliability.
I have had a variety of SCSI drives, starting with 2gb 10mb/sec SCSI. Most are not currently in use.
Well, I had my first serious IDE hard drive failure. I bought two 400gb ide seagate hard drives (on sale for $90 each with a 5 year warranty), and installed them in my computer. After a week or two, one of the hard drives failed in a novel fashion. If it wss plugged into the computer, the computer would not boot. If it was disconnected, the computer would boot. I contacted frys.com and they sent another drive, and they say they will pay for return shipping of the broken drive. This is a much better deal than sending it to seagate, as they would send me a refurburished drive. I don't think seagate makes junk; I think I was the victim of infant mortality
It is well known that electronics fail in a non-obvious fashion. When electronics is new, the failure rate is quite high. When electronics gets very old, the failure rate goes up. This makes a 'bathtub' failure curve. The new failures are known as 'infant mortality' and the old failures are considered to be the device wearing out. The military solution to this is to 'burn-in' new electronics at elevated temperatures and voltages for a time (often one week). The extra temperature and voltage will often cause marginal parts to fail. This greatly reduces infant mortality, but at a high cost. Since I paid $90 for the hard drive, and it worked fine when it was new, I conclude that the failure was due to infant mortality. So if you want a really reliable computer, run it for a few weeks. This is why I only use used hard drives in my file server...
I have had a variety of IDE drives. I have seen a few serious failures of drives that I have inherited. For all drives I have purchased, when there were problems, I could always copy most or all the files before the drive failed.
I usually run my main computer 24x7. It is very well cooled. See Keeping Everything cool for details.
I recently bought four 500gb sataII hard drives to use in a raid array. I bought them because they were inexpensive and had a 3 year warranty. Though they are labelled MAXTOR STM3500630AS drives, they are really Seagate 7200.10 series drives (Seagate recently purchased Maxtor). When I built the raid array, it failed immediately. I thought the problem was the linux driver for the sata II controller chip. I was wrong. One of the drives kept failing, leaving the array. For debugging, I booted windows xp and I was able to format each drive and copy stuff to it. XP didn't complain, so I thought it was a linux problem. Shows how weak the error reporting of XP is...
As hard drives get denser and more complex, the ability to monitor them has also increased. There is a standard called "SMART" which means Self-Monitoring Analysis and Reporting Technology. It reports all kinds of data about hard drive health. Thanks to the help of the linux-ide kernel folks, and me learning more about linux ide/sata, raid, and SMART monitoring, I found out that one drive was quite sick, and returned a bunch of errors of "READ DMA". This started after 57 hours of being powered up. The other drive returned several "READ VERIFY SECTOR" errors after 11 hours of being powered up. I was able to get these errors using "smartctl -a", which is a linux program, (ported to windows) that monitors SMART parameters. I send the drives back for repair/replacement. They were replaced with 750gb recertified hard drives.
Well, a few months after I got the two replacement drives for the broken maxtor drives I tried to build my new backup file server using four sata drives. Two were maxtor, one was the recertified Seagate 750gb drive from above, and the other was a 500gb Hitachi drive I had lying around. I built the software raid 5 array, but I kept on having errors with the files I wrote. I switched almost everything - the power supply, the processors, the motherboard, the memory, but still the errors persisted. It was very confusing tracking down the problem. I suspected samba, the network card, the controller card software, and lots of other stuff. I eventually broke apart the raid array, and tested each drive separatly. It turned out that three drives were perfect, and one wasn't. The fourth drive was one of the two Maxtor 500gb drives. Although there were no SMART errors, and it passed all of the Maxtor diagnostics, it still failed my test.
My test was to format the drive, and write sixteen big files. I happened to have dvd iso images for a linux distribution that I used for the files. Next, I generated a md5 checksum for the files and wrote it out. I then regenerated the md5 checksum, and it wasn't the same for several files. I am pretty sure that the bits didn't get switched on the disk, because when I generated more md5 checksums, they kept on changing. It was clear that I was getting disk read errors. There were between four and sixteen errors that I detected during reading of about 64 gigabits. Not a high error rate, but definitly a bad drive. I recommend running this test, or something like it on suspect hard drives. I got the replacement hard drive (a remanufactured 750gb), and I filled the disk with big files and ran the test a few times. No problems whatsoever. I am disappointed that none of the manufacturers tests showed this problem. On the other hand, they did replace the 3 failed drives with 750gb drives, so I got a total of 750gb for 'free'.
I have a Adaptec 2940UW SCSI controller. I bought it to replace the
classic Adaptec 2940, which came with my Gateway computer. It has
worked flawlessly for several years. A few days after I installed
my new Plextor it suddenly
stopped working. After several hours of thrashing around, trying
3 different SCSI cables, and several single devices to terminate
the narrow SCSI chain, I came to the realization that the narrow
SCSI chain part of the controller, but not the wide SCSI chain
part of the controller was fried. I don't have any idea how it
happened. After a few calls to Adaptec tech support, they
verified that it was a retail controller, with a 5 year warranty,
and they agreed to cross ship me a new controller. Needless to
say, the new controller worked flawlessly. I was totally shocked
that the controller could fry, and it has been my first hardware
failure. But the Adaptec 5 year warranty is really great, and
shows they are serious about standing behind their product.
I have a Sony 4mm DAT drive. There is now a newer one, the SDT-9000, which is faster and has a higher capacity (what else is new?). It is reasonably reliable. I have had to send it back once, since it broke in a way that I don't recall.
Well it broke again 12/27/97. Did a full backup flawlessly, but refused to eject the tape. The manual eject button does no good, and when I send a SCSI command to eject, I get an error return. I am most unimpressed. I have probably made 20-30 tapes, and it has failed twice. I will send it back to Sony, and no doubt pay plenty. This drive is still listed on their web page, and people are still selling it, so I would guess it is still being made. Hope Sony can get my backup tape out... I now do not recommend this drive at all under any circumstances due to poor reliability.
Well it broke again Dec-24-1999. Backed up some data, then got a buncha errors, timeouts, and the like. And it was with new tape. So I traded it in with New Image International. They told me that the Sony sdt-7000, sdt-9000, and sdt-10000 have 'soft' heads, and when they go, the cost to fix the heads is about what the drive is worth. They said the HP DAT-8 was better, so I got it. I will eventually get a DLT drive, when they get cheaper and I get more money.
I bought this in 2000. In 2007 it failed. The power button doesn't turn green when you press it. No idea why. I guess I got my money's worth since it lasted 7 years.
If you have comments or suggestions, Email me at turbo-www@weasel.com
Created with gnu emacs and template-toolkit, not some sissy HTML editor.
No Java or javascript needed to view my web pages. They
both have significant security issues.