Please register or login. There are 0 registered and 1215 anonymous users currently online. Current bandwidth usage: 326.30 kbit/s August 08 - 06:18pm EDT 
Hardware Analysis
Forums Product Prices

  Latest Topics 

More >>


  You Are Here: 
/ Forums / Hard Drives /

  How to decide before failure that hard disk needs to be replaced? 
 Date Written 
Hit Machine May 13, 2011, 04:42am EDT Report Abuse
Are there any parameters by which one can get to know that a hard disk performance is not up to the mark and needs to be changed before the actual failure happens?

Want to enjoy fewer advertisements and more features? Click here to become a Hardware Analysis registered user.
Juan Pena May 13, 2011, 06:24am EDT Report Abuse
>> Re: How to decide before failure that hard disk needs to be replaced?
there are easy things you can do to verify its integrity. run check disk, scan disk from Proporties - Tools; you can also download some SMART program, and even check in t he manufacturer's website for any program for said purpose.

If you get any 'delayed writing..', cannot save a file on the hd, these are also signs something is not right.

sooner or later, all hdds fail so better be prepared.

Good luck, Juan.

Hit Machine May 13, 2011, 09:01am EDT Report Abuse
>> Re: How to decide before failure that hard disk needs to be replaced?
I have used Active @ Hard Disk Monitor and it calculates various parameters at its own. eg. My hard disk has throughput performance of 52%. The software says status of hard disk is Ok. But I feel that 52% is quite low. Similarly other parameters some come 60% and some perfectly fine.

I don't know should I replace the hard disk for better performance or not?

john albrich May 13, 2011, 09:40am EDT Report Abuse
>> Re: How to decide before failure that hard disk needs to be replaced?

There's Passmark's "Disk Checkup", which monitor's disk degradation over time and attempts to provide a "time to failure" date. While the date may not be accurate, it has successfully pointed out a few drives that soon failed/became unreliable.

I also like Speedfan's method of comparing drives to an ongoing increasing ONLINE database of similar disk drives. Again, not necessarily a date, but it does ID drives that should be closely monitored or just kept backed-up (or replaced if you want to be extra extra cautious). Even if you decide not to use speedfan to monitor your system parms, it would still be useful for week checking and interpretation of drive SMART parms.

I talked about both of these programs in a different recent thread. If I find it, I'll post a link to it...otherwise, you might try searching HWA's search bar for my last name and "checkup".

PassMark DiskCheckup


I think they're both also available from,, and

Hit Machine May 13, 2011, 03:36pm EDT Report Abuse
>> Re: How to decide before failure that hard disk needs to be replaced?
SpeedFan is really a nice and comprehensive software.


john albrich May 13, 2011, 05:56pm EDT Report Abuse
>> Re: How to decide before failure that hard disk needs to be replaced?
Manik Sood said:
...I don't know should I replace the hard disk for better performance or not?

HDtune and HDspeed are decent programs for assessing HDD performance. *

Both programs can easily show you the performance impact from such things as excessive read errors (which can degrade as the drive ages), USB v SATA v IDE connection, 5400rpm v 7200rpm, HOW you format a partition (e.g. whether you format using the default 4KB cluster size or use 16KB or 32KB), WHERE on the disk platter the head assembly is currently located (for example HDD sequential read performance might be 120MB/sec in the first 5% of the platter but degrades to 30MB/sec by the time the disk is filled up and the head is near the other end of the platter...and that is the way it is SUPPOSED to work...everything might be fine even though it might look suspicious), and so on.

By the way, a common situation might be...
Because the drive head assembly's physical location on the platter can make a very fast disk seem slower than an actually much slower disk, the user could misinterpret the results and make a mistake by concluding there is something wrong with the higher performance disk drive when there is NOTHING WRONG. (If I didn't explain this well enough, please let me know and I'll try to clarify).

I've started taking performance 'snapshots' when I first receive a drive, and can compare the initial performance to the drive's performance as it ages.

Warning!-HDspeed's WRITE test is a destructive low-level test. You will lose data on the tested partition and (at the last version I used) you have to reformat the partition after using HDspeed's WRITE test. HDtune uses a high-level (Windows-file based) test and is thus non-destructive. Both HDtune and HDspeed's default tests are a non-destructive READ test. And, as far as I'm concerned, READ tests are probably all you really need in 95% of situations when you are investigating a possible disk drive performance degradation problem.

The reason I like HDspeed for SOME testing is that you can simply instruct the program WHERE on the partition you want the test to physically start (e.g. at the 50% point, 30% point, etc). This can save a lot of time if you want to test specific sections of the disk. Otherwise for HDtune, you have to pre-create multiple partitions of sizes that are consistent with the physical locations you want testing to occur if you want easily to know WHERE on the disk you are testing.

There is some other point I was going to make, but the painkillers I'm currently taking fuzzed it up before I could type it in. If I remember, I'll edit this post...
Edit to it is--
HDtune is particularly good for visually showing when a disk has physical problems: e.g. read errors, head/track alignment, "high fly writes", and so on. These might show up in the performance graph as sudden somewhat consistly located dips in the performance graph's continuous line, whereas the impact of tracking-related errors might best show up in the "Access Time" scatter plot and higher than expected access time value (e.g. it might be quite a bit longer than spec or longer that it was 2 months ago). The appearance of the scatter plot also matters. A narrow well-defined pattern shows consistent drive operation. If the scatter plot is however "all over the place" and/or if some "loner" datapoints wildly diverge from the "band", you are likely seeing tracking and/or HDA electronics issues. Note: such can be aggravated by drive temperatures.



  Topic Tools 
RSS UpdatesRSS Updates

  Related Articles 

A weekly newsletter featuring an editorial and a roundup of the latest articles, news and other interesting topics.

Please enter your email address below and click Subscribe.