• Content Count

  • Joined

  • Last visited

Everything posted by pandabear

  1. pandabear

    About swap file position, new theory

    HD is faster in the outer region (beginning of a drive) due to higher Bit Per Inch ratio, while your theory may reduce seek, outer region have less track per MB and thus will also have a smaller seek time. The trade off as you mentioned, is whether to use this faster region for your program or swap. It is impossible to tell for your application without doing a benchmark on it. That is a very big assumption you made here. It is hard to say now a days which part of the program is executable and which part is data, and by using your swap to chase your program/data is not a good idea, expecially if you lower your BPI by moving it to the middle of the disk instead of the Outer diameter. The better way to do it is to put your swap at the beginning (OD) then use a good defrag tool and adjust its setting so the program you need to swap alot will be near the swap, this is the more efficient way.
  2. pandabear

    performance drop over time?

    I am suspecting that is due to the improvement on the new drive instead of the old drive getting slower. Maybe a new generation of drive is larger, more buffer space, has improved ASIC and cache algorithm, or etc that makes your experience faster. I can imagine seek getting slightly slower or the startup head load/unload getting slower, but no way would it be 1/2 the original without failing.
  3. I am a graduate student in UCI. Recently the admins contacted me to take over two 25% TA positions for another student that quit. Based on previous experience and common practices. I got paid and my tuition ($3000) will be paid for. Now here is the part, they never send me the contract and when I realize they haven't paid for my tuition, I asked and they avoided my question until last week, 1/2 way over the quarter. That they won't pay my tuition because I don't start on the first week. But I am just taking over someone else's position and they refunded their tuition back to the school, so IMO they should at least pro-rated it to the amount of time I have served. They told me the decision is final and refuse to discuss with me. I got fed up and quit. Now here is the even crazier part. Before I quit I had already signed a contract with them for next quarter TAing. They emailed me today saying that if I don't get back to this quarter's TA positions they will void the contract for next quarter. I am completely fed up by their bait and switch tatic that they try to shuff down my throat. Even now they still haven't send me the contract of this quarter that I signed. I may need legal help as I am not sure what other things they may do to me or other student. One of my classmate got a lower grade with perfect work because the professor doesn't like him, and another student got kicked out of his TA position by a staff member regardless of his professor's numerious protest, because the staff want this TA position assigned to his/her favorite student that works for his lab. This is totally politics and serve the school or student no good. If you have any idea on what I should do, please email me and help.
  4. pandabear

    RMA back to WD is purged data safe?

    Is there a way to know whether a recovered HD is soft or hard recovered? Do you see any difference in terms of reliability between the 2?
  5. Lexwalker, if you don't want to be treated like a spammer you have to stop acting like one. Although I did feel irriateted by your former post against maxtor and feel that it is understandable. When you keep trashing one company in every single thread with blanket rant and suggesting everyone to read one data recovery company's website as a reference, it makes the experience here intolerable. So please, if you don't want to be called a spammer, stop acting like one. Also, you will find that many storage review visitors to be in the storage industry insider, from QA to engineers to OEM buyers to IT professionals that deals with between hundreds and millions drives a quarter on a regular basis. You will also find people that have worked in many company through out their life in different rival companys (i.e. jumping between WDC, Seagate, IBM, Quantum, and Maxtor) and between different part of the harddrive food chain (from Dell, EMC, Veritas, Hitachi, ST, Marvell, Agere, ReadRite, TDK, Nidec, etc). Once you stay long enough you will be amazed, on topics like how one vendor's chipset has a bug, how another vendor's RAID is done in software instead of hardware, or how one wierd symptom is fixable by which firmware version. There are many things to learn here. Back to that particular data recovery company. I am making an assumption that you have no affiliation with them, so I am going to speak from my mind about why many of us would rather throw the drive away than wasting $ on them 1. Making a blanket statement about why one brand is always going to fail is like saying every one who eat bread will die one day. This is pointless. 2. Making a blanket statement about all drive in a particular model is recoverable is asking for some. This shows how innocent/ignorant this recovery firm is in the industry and how much you can trust them. 3. The way they "fix" a DM9 by soldering a 5V line to one phase of the circuit is lucky at best and dangerous at worst. If they are not at least replacing the PCB while trying to recover the data, I won't be confident with them. 4. No offsense, but having someone hardsell it in every single thread is the biggest turn off for serious customer.
  6. Yes, DM8 was well known for having some problem to begin with. I person had one 30gb unit died on me after 1 month of use, 247 in a bittorrent/kazaa client with T1 connection (don't ask). It is not up to my standard. However, the replacement come immediately after (Maxtor do have good and quick warranty, the best I have seen out of all warranty I used in the past) and is performing the same role well for more than 1.5 year. It is also much cooler than the one I replaced (33C vs 46C), so I am confident that it is good for at least I need to upgrade again (knock on wood). I am not the only one getting DM9, check out Unofficial Dell PowerEdge SC420 FAQ *BETA* and you will see they are using Calypso 80GB SATA (Calypso is DM9), and no user seems to report any HD failures. For this reason I am fairly confident. BTW, it is almost impossible to directly calculate how much one company spend on RMA to determine their reliability. Those % you listed is a percentage of sales, if they lost sales, have high cost of RMA (i.e. located in singapore vs china), longer warranty period, tough warranty policy (any one remember eMachine denying warranty on individual part, you have to send in the whole machine ?), etc can make it very different. FYI seagate is now manufactured in china, so they are going to cost much lower than Maxtor which is in Singapore and WD which is in Malay. Same goes for MTBF, it is only an indicator of Infant Motality (how many unit is Dead right off the bat), because they test say 5000 units for 1000 hours each, then call it MTBF 5million hours. It doesn't tell you how long it is going to last over its life. All in all the only good indicator is the reliability data base here and each manufacture's internal data (won't have access).
  7. I think you have already made your point in your other posts, no need to flood the rest of the forum with this.
  8. You forgot about the topic which started this thread which was: According to the DM8s suffers more than the flyheight adjustment issue. It also stated DM9 problems. 197507[/snapback] Chill dude, I understand you were frustrated with the maxtor drive you have to deal with in your last job that you quit. I think you have already made your point and sorry to feel your lost. However, your quote about 90% failure rate in Maxtor drive is totally out of any logical assumption. No company can survive for more than 1 quarter with that kind of result. The fact that DELL is as of Dec 2004 still using Maxtor (DM9 in particular) in many of their line is a prove that their quality is at least on par with the rest. From my rough memory they have RMA rate at most 1% higher than the rest (everyone is between 7-9%), I am not 100% sure about the number but they are very close, nothing significant. Yes I have read your link PDF, it is complete BS to me as a point of view from an electrical and manufacturing engineer. A friend of mine from Maxtor's QA told me they had a problem in the past during manufacturing, they pull the PCB off a drive's HDA before they completely spin down, and cause the back EMF to charge up too much voltage and weaken some power electronics or cause instant failure. I think this is the "common" electronics failure that was mentioned in link. They found and fix it within one quarter, so I don't think all DM9 that are in production recently (in the last 1 year) is affected by this. BTW, one friend of mine always have WD failed on him and he vow never buy any WD again. When I take a look at his case i found what happen, 5 drive stack on top of each other in 3.5 bay without cooling, 5 drives. Any brand and model will fail in that environment. You may want to take a look at why 90% of your company's drive fail, because it may be a bad batch (refurb without being labeled), bad distributor that don't know how to handle drives, or poor system design (no cooling/ power supply problem/ people kicking and droping their PC all the time).
  9. pandabear

    'lemon' 6Y080L0 sata

    Check what UDMA mode you are running in and whether you can swap the IDE cable, I had the same problem in the past running 40pin cable and running in PIO mode because I want to save a few penny. Check that one first.
  10. pandabear

    'lemon' 6Y080L0 sata

    Well, if you look at the main site of dataclinic, they have list of fujitsu and western digital as well. Beside, I don't think quoting a data recovery company for reliability of HD in the past is a good indicator, they may only show what is their experties in. from what i know the D740, 540X, and Fireball 3 has no reliability problem what so ever, so why would they put it there? (maybe they were a maxtor warranty contractor in the past).
  11. DELL only stopped using the DM8 due to flyheight adjustment issue. They are still using DM9 that I don't think has any major problem. I just got an SC420 that has a DM9 in it, no problem and works great so far. From what I know Maxtor did got a fix in their DM8, but they decide to cancel that line in the near future because of the already bad rep and the small capacity that they can't charge a lot for it.
  12. pandabear

    Hitachi vs. Maxtor...which do you recommend?

    2 weeks turn around just to change the brand of HD you get and the hazzle of reinstalling everything? No, thats not an option. I am sure it would be cheaper to just buy a few spare elsewhere than to replace a drive that may not have a problem, with another drive that also may not have a problem.
  13. pandabear

    Hitachi vs. Maxtor...which do you recommend?

    I think for most people here they are taking the minor differences too seriously. IMO you are fine with either drive. If you feel more comfortable with IBM/Hitachi in the past then stay with it, or vise versa for Maxtor. In general the Maxline series in Maxtor do get more attention to detail in terms of the manufacturing test process and are the "prime" drives in the bin. This is similar to how Intel use the same wafer but test the best for the higher speed one, then the lesser one for the lower end products. No industry rumors suggest reliability problems on either drives (subjective opinion may be different depends on user bias, as usual). I think you get more variation by who you get your HD from than which one you get. Like if your vendor is handling them correctly, etc. If they don't know how to ship them or drop them, either drive may have problem very soon.
  14. Yeah, I do agree that 5 year warranty is really cutting into their balance sheet. However, since they make them in China now it is much cheaper and they can afford to do so. They do have a slower seek and this makes them in theory more reliable.
  15. pandabear

    5 1/2" Hard drives.

    This is similar to why our cars have 6k to 9k redline. Sure it is possible to build something faster but at what cost? If you end up being more expensive than using smaller disk but more disk, you already lose the battle. The last generation of 5.25 inch is Quantum bigfoot, suffering reliability problem and spin very slow (3600 or 4200rpm) only.
  16. Today's harddrive design rely heavily on ECC to get the optimal density out of it. No drive manufacture can build a drive without using ECC to compensate for the error, but the strength of ECC is very strong and I wouldn't worry too much about it. Just an example, my Seagate 7200.7 has the following smart result, kind of scary but it is working perfectly fine: SMART & Simple for Windows NT/2000/XP V1.01 Copyright 2001-2003 Opened Drive \\.\c: . . SMART Enabled : Yes Model Number : ST340014A Firmware Version : 3.16 Serial Number : 3JX7PW0P Drive Size : 40.000 GB ID Attribute Type Threshold Value Worst Raw Status ---- -------------------------- ----- --------- ----- ----- ---------- -------- [01] Raw Read Error Rate Prefailure 6 66 62 182578302 OK [03] Spin Up Time Prefailure 0 98 97 0 OK [04] Start/Stop Count Advisory 20 100 100 8 OK [05] Reallocated Sector Count Prefailure 36 100 100 0 OK [07] Seek Error Rate Prefailure 30 79 60 95537422 OK [09] Power On Hours Count Advisory 0 98 98 2302 OK [0A] Spin Retry Count Prefailure 97 100 100 0 OK [0C] Power Cycle Count Advisory 20 100 100 31 OK [C2] Drive Temperature Advisory 0 45 48 45 OK [C3] ECC Corrected Count Advisory 0 66 61 182578302 OK [C5] Pending Sector Count Advisory 0 100 100 0 OK [C6] UnCorrectable Sector Count Advisory 0 100 100 0 OK [C7] CRC Error Count Advisory 0 200 200 0 OK [C8] Write Error Rate Advisory 0 100 253 0 OK [CA] Unknown SMART Attribute Advisory 0 100 253 0 OK
  17. pandabear

    Latest Biblical Intel On World Events-AsianTsunami

    So you just sign up a membership to post this? with 4 post total?
  18. Just like others said, with UDMA you hardly have to worry about CPU utilization anymore. On my Athlon 1600 with DiamondMax 8 30gb and a LiteOn 52X, I have no problem continuously burning at maximum speed with only 5% CPU utilization. The key is to make sure UDMA is enabled for the CD burner instead of the harddrive. I heard that if you got read error windows will automatically tune it down to PIO mode which is the original ATA mode, very slow. It seems like the current generation hard drive the interface is no longer about speed, but more about whether you are paying to get reliability, over-engineering, long warranty. Now the SCSI is fast because they have better ASIC (dual processor and 2x the amount of on chip cache), faster motor, faster actuator, instead of the good of the interface.
  19. They came out in 2003, so it is not your fault. Seems like you can get it on ebay new for $50-60 each shipped now. =)
  20. pandabear

    Quick Format Harmful to Hard Disk?

    You get the same "wear" on the FAT table, inodes, or whatever partition type's head portion, but full format will "wear" more on the "content/data" part of the partition. Your friend's comment doesn't make sense on a HD, since ECC, sector reallocation, calibration is done internally on the drive, it doesn't make any sense that any user read/write (what full format really is nowadays) will improve/fix/detect anything except perform a complete read/write on the entire partition. It does make sense for a floppy that it will verify / write blank the entire surface so you will know if there is any bad sector. But for HD you better run a smart diag software or any diag software that comes with the HD.
  21. pandabear

    Apple mini, what HDD inside?

    Yeah, so far Maxtor still hasn't got a 2.5" yet. But from what I heard they will have one out very soon (within a few months). Don't know what it is going to look like yet.
  22. Good news for you, Hitachi has an Endurastar that is 2.5" ATA5 that has 20gb and -40 to 85C op range. Seems to plug right in for you.
  23. Also, make sure you check if the ambient pressure change as well, HD doesn't like pressure to go up or down too significantly. Remember, PV=nRT and such a huge temp change may cause pressure to change, and these 2 factor along could cause the head to fly much different than designed.
  24. Under that kind of ambient temperature and exp it is being a 2.5" size laptop HD, it could be very true that FDB is failing. You are talking the fluid being 30C hotter than design/spec so it may be out of spec and doesn't last as long. You need to remember that it is 1-4 hours in those condition per day, not just a flash to those temperature, fluid doesn't like that too much. Ball bearing may last longer in those application as they can take more abuse, but it is hard to imagine any drive design for that kind of spec. If you are designing new enclosure, maybe try to get better cooling or relocate the drive elsewhere. If it is not possible then your only choice maybe to switch to solid state (flash). It may be a lot of $ to redesign, but you can pay now (redesign) or you can pay later (drive replacement material and labor, tool downtime, losing business due to bad reputation).
  25. pandabear

    Maxtor drives: Are there problems?

    Yes and no. Diamond Max 8 (the thinner profile one with single head, single platter, ranging from 20-80GB) had earlier problem with fly height adjustment. It was a problem in the early batch that causes very high failure rate. A fix in firmware resolve this and they developed temperature based fly height adjustment and its reliability is now back on track for the later unit. I had one failed after 1 month of use and the replacement work like a champ 247 for over a year. DELL drop this account because of it and they ask Maxtor to requalify with 2500 hour test instead of 1000 hour like everyone else. My new SC420 came with a Diamond Max 9 and it works like a champ. PandaBear