polling files from disk limit hard drive life expectancy?

[Aarstad]

Hey guys.

I've come to a realisation that a computer will turn off the hard drive if it's not in use. I've also made a software that polls the hard drive for a file every 30 second or so, which got me thinking.. will this severely dampen the lifetime of the hard drive?

How often does a software normally poll the hard drive, or use the hard drive? I am uncertain how long it takes for windows for telling the hard drive to go to sleep, but a poll every 30 second would keep that drive active pretty much day and night, right?

The thing is that I am looking for a file that suddenly appears with a log, and when that happens, the script will notice it and upload it to a mainframe I've set up so we can read it. This is to ease the process of checking logs instead of manually logging in everytime to check it.

[FireFox]

Hi,

You can tell Windows to keep the drive started. End of the problem.

Edit: I don't understand the last part with the logs.

Br, FireFox.

Edited August 6, 2013 by FireFox

[Edano]

you can tell windows to never shutdown harddrives and/or monitor. energy settings i think.

[AdmiralAlkex]

HDD life expectancy doesn't seem to be an exact science. I think I've read both about heat and on-off cycles as the main reason they wear out. If you can keep a decent temperature with them always on, I would just let them run for the heck of it.

Also check the >FileSystemMonitor UDF, maybe that doesn't spin up your drive (never really used it myself).

Or if you decide to continue polling, think if you really need it to be as often as 30 seconds.

[BrewManNH]

I used to seed torrents using bittorrent and I've killed a lot of hard drives that way, 3 in one year at one time. Newer hard drives don't seem to be as robust as they used to be, so running a hard drive constantly will definitely limit its life expectancy.l

[Edano]

I used to seed torrents using bittorrent and I've killed a lot of hard drives that way, 3 in one year at one time. Newer hard drives don't seem to be as robust as they used to be, so running a hard drive constantly will definitely limit its life expectancy.l

.

that is true. mostly the electronic fails, and it's nearly impossible to find the suitable chipset board for a distinct drive. you have to find another drive with a similar serial number, when it's older than half a year maybe you're lucky on ebay.

[ripdad]

@BrewManNH <-- Hard Drive Tormenter 

@Aarstad - What you are proposing, 30 second writes, will have little effect.

Although, there is a way within Windows to set your Power Scheme to "Always On".

Heat, shock, heavy read/writes every second, hard crashes are all enemies of a hard drive.

A busy operating system, such as constant virus scans, backups, logging, and not to mention ... seeding torrents.

And, among other things like quality of materials, head assembly, thickness of the magnetic surface and engineering.

Lets not forget an overheated CPU. It can do damage to a hard drive in the form of bad sectors and data gets strowed everywhere.

I remember when most hard drives would outlast the life of the computer.

Hmm, where did those days go?

[robertcollier4]

If you're reading the same piece of data repeatedly every 30 seconds and there is no other disk activity going on - then I assume that Windows filesystem cache would cache that read so that repeated reads would be served from RAM instead of from disk.

Sysinternals Diskmon will show you activity on the disk.

http://technet.microsoft.com/en-us/sysinternals/bb896646.aspx

Sysinternals Process Monitor will show you activity to the filesystem.

http://technet.microsoft.com/en-in/sysinternals/bb896645.aspx

You can run both Diskmon and ProcMon side by side. And you will see that all the filesystem requests that show up in ProcMon don't show up DiskMon. When file system activity shows up in ProcMon but not in DiskMon that means that the file got served from the filesystem cache in RAM.

If you want to try to do the operation fully in RAM, You can also:

1. Extend your filesystem cache using a tool such as PrimoCache, eBoostr, or Superspeed Supercache. Windows filesystem caches on a file level basis - these additional caches will then also add another cache layer on top of that to cache reads and writes on a block level basis.

2. Make everything operate off of a RAMDisk by creating a RAMDisk and then junction the directory where the log file will be created to a directory on the RAMDisk.

Edited August 7, 2013 by robertcollier4

[Edano]

i think the best way to kill a harddrive is defragmenting it too often sometimes you can hear them crying.

Edited August 7, 2013 by Edano

[ripdad]

@Edano - Not unless you defrag everyday.

There is no reason to defrag your hard drive more than once a month, at most.

I personally, defrag my hard drives about every 6 months. They all have over 40,000 hours on them. (Western Digitals)

@AdmiralAlkex - You're right about  "doesn't seem to be an exact science" about them.

But overheating, in my opinion, is at the top of my list. That goes for the CPU also.