Well well well… our post of a reader’s conversation with an IBM tech has certainly reverberated across the community far more than we anticipated and intended. Feedback on the issue has poured in from all sides.Allow me to take a moment to point out that StorageReview.com historically has rarely if ever commented directly on reliability issues. The fact is that we simply don’t possess the resources to accurately investigate reliability reports and that we don’t want relatively vague comments on reliability to dilute other more exhaustive and precise measures… the care that goes into our performance suites, for example. Case in point, we haven’t had anything to say one way or another on the alleged problems of the Deskstar 75GXP; though in the past we’ve been inundated with traffic from large sites citing “SR comments,” they invariably were actually links to various participants in our discussion forums.
That said, it’s important to note that March 6th’s headline may have stepped beyond SR’s self-imposed bounds. We regard the front page as sort of an informal way to publish the latest in news, reviews, etc. Blurbs contained within are not to be regarded as “canon” in any way. The posted transcript summarized a conversation between a long-term SR community member and an IBM tech. For obvious reasons, both persons remain anonymous.
After witnessing the fallout of the post, however, it became clear that we shouldn’t be so “fast and loose” with FP blurbs and resolved to discontinue what was becoming “rolling commentary” of sorts on the 120GXP. With some merit, IBM points out that we not only allowed one side to get the “last word,” we allowed that side to also have the only word.
IBM claims that, contrary to what we stated on the 6th, they did get back to us in a timely manner via e-mail. We’ve run into mail problems several times in the past so we have little reason not to believe them. A spokeswoman has this to say:
Regarding the conversation between the storagereview reader and the unnamed source, IBM does not know who that source is and does not support or confirm the authenticity of those statements. Our position is that users can certainly run the drive more than eight hours a day.
She also adds the following points:
The 333 power-on hours (POH) defined in the 120GXP data sheet is not a new spec for our GXP drives; it is consistent with what we've held our desktop drives to in previous generation drives. The 333 power-on spec is not an indication of a maximum number of power-on hours or limitation of the Deskstar 120GXP.Our specifications indicate that the 333 power-on hours per month represent typical desktop PC usage. This assumes an 11-hour day based on a 30 day month. Users can and have successfully run the drive more than 11 hours a day and 333 hours per month; the drives have been used successfully in 24x7 environments.
IBM stands by the 3-year warranty for the 120GXP. Power-on hours will not be a determining factor in negating the warranty.
Since SR itself expressed hesitation on 120GXP server usage, I asked this question: What's IBM's official position in using Deskstar products in servers that would remain on 24/7? Though the warranty would be fully honored, is server utilization encouraged, discouraged, or neither?
IBM’s response:
Use of the drive 24x7 will not affect or void the warranty. IBM stands by our three-year warranty for the 120GXP. The drive can and has been used successfully in 24x7 operations. Users should not be concerned about running the drive in 24x7 environments beyond concerns typically associated with any drive usage, such as operating temperature, vibration, duty cycle, etc.
This is IBM's side of the issue. Ultimate conclusions when it comes to these nebulous rebiliabity issues are, as always, left to the reader.