glug

Member
  • Content Count

    382
  • Joined

  • Last visited

Community Reputation

0 Neutral

About glug

  • Rank
    Member
  1. Software RAID or fakeRAID doesn't require any special kind of drive, I didn't think hardware RAID did either. It won't remove the drive for taking a minute to remap, instead it will basically hang the system for that minute. As far as error recovery I guess it depends on whether the drive thinks it "fixed" the error with remapping or reported the error to the controller? If the drive reports the read error, it should recover it from a mirror in RAID1 even if its software RAID. There are exceptions to this and it is of course implementation dependent. Using 2 fusion I/O drives in software RAID1, rather than letting you use the other drive when one of them died Windows refused to mount the array entirely - had to take the good drive out of the mirror set and just use it as a normal drive until a new card was delivered to replace the dead one to be re-mirrored to. You should be able to use the SSDs in whatever type of RAID you want, however I would recommend against using RAID5 as you double the # of writes to maintain parity without getting the safety level of RAID 1. Also, if the RAID5 is badly implemented and always uses the same drive for parity rather than rotating it, the parity drive will get wear-leveled faster than the others.
  2. I've seen way too many Maxtor drive failures over the years to ever trust them. While working at an ISP in '94-'97, I saw 2 Fireballs live up to their name, as in they literally burst into flames, on 2 separate occasions. Both of these were used in the news (NNTP) server, which chewed through quite a lot of drives -- I think it is harder on them than most stress tests. One of my friends in college had his Maxtor 540MB drive start smoking, but nothing as spectacular as the fireballs. I've also seen a ridiculously high failure rate in a company that had mostly Conner drives shortly before they got bought by Seagate. It seems like they've made a habit of buying companies with shoddy manufacturing quality over the years.
  3. Also useful for people living in New York City. Con Edison is worse than any 3rd world power grid.
  4. RAID 1 is going to be faster than RAID 0 for any game where disk speed is an issue, other than during installation time, since game disk access is almost entirely reading, not writing, and it's not all that sequential.
  5. glug

    A drive to AVOID

    Could you provide the full model code with the suffix part of (like WD2000JD-xxxxxx) for both of these drive versions? Just interesting to know for statistics... The original drives were all WD2000JD-00GBB0. Replacements are WD2000JD-00HBB0 (2x) and WD2000JD-00FYB0. (Haven't RMA'ed 4th drive yet). I assumed the replacements were using the newer 100GB platters as they seem slightly quicker than the GBB0 drives did, but I just looked these #'s up in another thread and it says they are all 80GB/platter?
  6. glug

    Extreme temp Hard drive?

    I think those reports are WAY too conservative. I know that in hot & humid summer weather, the temperature in the passenger compartment of the car gets higher than that while the car is sitting in a parking lot. You also have to take into account that the ambient temperature for the drive is the temperature inside the PC case, which is going to be a bit warmer than the rest of the trunk. It's also going to be too hot for the CPU and other components, and I would not be surprised at all if the plastic pieces on the motherboard literally melt away.
  7. glug

    A drive to AVOID

    Funny you should mention a laser printer, I actually do have one that I rarely use plugged into one of the 2 non-battery backed outlets on the UPS. It is a small one though, LJ1012, which according to specs uses 250W when printing (2W when idle). I used it the night before the drive died, for the first time in a month. The computer, monitor, router, cable modem, & scanner hooked up to the UPS use about 250W normally, sometimes up to low 300s depending what I'm doing. The printer should be moved to a separate outlet?
  8. glug

    A drive to AVOID

    My case has 2 110mm fans in addition to the PSU fans, with one of the fans in front of the drive bays, so the cooling is pretty good. The first 2 failures you could hear the drives make noises as they failed to read. The 2 failures this year, the BIOS post would hang (first one indefinately, 2nd one for about a minute) after listing them if the drives were connected during startup (and it would make Windows hang all over the place as well if you connected them after BIOS startup). The first of those 2 failures, the drive was actually listed twice in the BIOS startup screen. The 2nd one it would half the time not be listed at all.
  9. glug

    A drive to AVOID

    To say that the power in Astoria is marginal is the understatement of the year. Beginning in April, as temperatures creeped into the 50s & 60s, the UPS would switch to battery from brownouts below 90V practically every other day during the evening hours when people started getting home from work. The reason there were such an incredible number of electrical fires in Astoria is no doubt because the brownouts they had every day were causing everything to pull more current. During the 2 weeks before the blackouts, I had to switch the computer off several times because the battery was getting drained too low. The night before the blackout the electric outlet voltage dropped into the 60s. During the blackout, I came back to my apartment and measured the voltage at 32V. Since it wasn't 0, even though that won't power anything and in fact damages a lot of equipment, ConEd doesn't count that as a blackout. If they reported the REAL numbers it would have been over 1 million people without power for more than a week, not 100,000. Since the blackout, what they've been doing I guess to keep the brownouts from being as severe is that during periods of lower power usage, they are actually sending around 130V. ConEd's statements about having the "most reliable power in the country" are a sick joke. I've had 2 phones killed while charging from the bad power also.
  10. I purchased 4 WD 200GB SATA drives back in 2004. 1 of the original 4 failed after about 7 months A 2nd one failed a few months after that. This year, the last 2 of the original 4 drives have failed, with the final drive death occuring this morning. I should note on these last 2, that the power input is a likely culprit, as I moved to Astoria before those failures. Let's just say that Con Edison's electrical grid quality leaves a lot to be desired, and even though the system is hooked up to an APC BackUPS 1500RS, I had the motherboard die 6 weeks ago with an imploded capacitor. All of these failures have been on Sundays, I strongly suspect something happens with the power grid weekly to cause this. The original 4 drives were all 3 platter versions. The 2 platter replacements have survived so far (knock on wood). Time to start reinstalling software. Again (uggh).
  11. According to the Department of Homeland Security, whose job it is to instill fear in all Americans, sweating on those 140F+ platforms is a sign you are a suicide bomber! http://www.newsday.com/news/local/wire/new...egion-apnewyork
  12. Before Bush is elected a 3rd time, he will have to be elected a FIRST time. That hasn't happened yet, the 2004 election was a sham just like the 2000 one.
  13. Ugh, never finished the upgrade (crashed more than once while attempting it), now the drive is officially dead and makes loud scratching noises when powered. Guess I'll be running on RAID0 and backing up while getting an RMA.
  14. ok, think I got it. Booted with the drive disconnected, connecting it after the array is detected. Converting the RAID0 array to RAID10 as we speak.