Swapping drives ?

 .

Is there any Dell supporting documents for number of times you can swap a drive. Some do not belive there is limit. I need some documentation to show the limited use of the swapping bays.

Thanks.

There is noI technical limit. (I suppose you could start to think about how many billions if times it would take to physically wear out the connection material.)

I have a vague recollection that in the early days the sata mechanical specification had some minimum limit on the number of insertions (connections and disconnections) of the sata connector itself and it wasn't in even in the 1000's. Maybe to today's connectors are constructed much more robustly so its less of an issue but who knows..

However, Dell should know how many insertions and removals their HDD sleds and HDD SATA/SAS connectors are designed to cope with.

Even if there is a theoretical limit, I doubt it is even practical to reach it. Its less than 20003 (should have read 2000) if you swapped it every day for the high end of the normal lifespan of a server. If you are engaged in daily swapping, your methods and strategy should be reviewed for improvement.

Sorry, theflash, your 20003 number confuses me.

If the insertion limit is indeed <1000, then that's <19 years (1000/52) worth of removing a disk drive once a week from a specific slot within your raid enclosure. This limit could be as low as <100 which equates to <2 years (100/52) worth of weekly swapping before a connector reliability issues causes problems. Obviously swapping a drive daily reduces these numbers to 3 years or 3 months respectively.

Such an approach as OP suggested could be workable from an equipment perspective at the upper limit of 1000 for weekly swapping a drive but unworkable at the lower limit of 100. Thus the importance of the OP's question on whether Dell has any documents defining the insertion limit/reliability for it's enclosure and disk drives connectors.

The issue of whether this is good practice is secondary to this question of insertion limit. Where within the Dell bureaucracy could one ask for such a document? Answering this would help the OP convince his cohorts to avoid such a backup mechanism (if i read between the lines correctly).

From a secondary perspective, using raid for ones boot drives and swapping drives as a backup mechanism may not be ideal but it's got to be better than no backup at all (if indeed insertion limits are high). But is it really workable? I'm not skilled with RAID so don't even know where the issues may lie.

LOL ... yeah, should have said 2000 (not sure how the '3' snuck in there) - so we can dismiss your math lesson :)

Only if you are daily swapping would you ever approach such a limit in the server's lifetime, and regardless of what the limit might be, it is poor practice and something else should be implemented if that is the best they have come up with.  You will pretty much be guaranteed to need to recover your RAID array(s) with such a practice - it is only a matter of time before you screw things up this way.  I disagree that it is secondary to the question - it is poor practice, making the limit (if one exists) irrelevant.  I'd be suprised if Dell would even have such documentation - engineering would be the only place to have such documentation, if it exists.  The way to dissuade the inexperienced/uninformed from implementing this bad plan (if this is indeed the purpose) is to dissuade based on the risk to the existing data, not based on the number of theoretical insertions.  The answer to them should be it is not possible; don't confuse them with maybe/could/should; simplify it for them.  If the techs are not able to deal with the eventual problems, it should not be attempted; if they know what they are doing with recovery and the server services and data are not critical, then I suppose they can go with it.

Even swapping external USB drives would be preferable - they are cheap and will be far more reliable - and safe - for the server.

Yeah your correct, it's poor practice and RAID (backup hacks) are not a replacement to a good backup and recovery practice.

Well this has certainly generated some discussion. I have tested it and it does work. I understand some of the issues why not to do it. For what its worth I have tried it and it does work. I have found that usb drives are sometimes  flakie.

Yes, of course it works ... pull the drive, insert new one ... but doing this with needless repetiveness is not a designed use of the controller.  It is only a matter of time before something goes wrong - I've seen way too many problems stem from people trying to do this.

I agree USB drives can sometimes be flaky, which is why I said "even" in my statement that 'even' USB drives would be better.