We are hosting few servers in datacenter locally and few of the hard drive started to give problem. Is this problem with datacenter temperature or hard drive issue? Any suggestions would be appreciated.
Sound like a power quality problem. Little voltage surges and spikes that originate from inside your facility often cause hard drive failures. These little surges and spikes are like repetive stress injuries for your equipment. They are not large enough to cause dramatic, immediate failures, but after a few hits they cause equipment to fail prematurely. Get a good double conversion UPS and then check your grounding.
Sound like a power quality problem. Little voltage surges and spikes that originate from inside your facility often cause hard drive failures. These little surges and spikes are like repetive stress injuries for your equipment. They are not large enough to cause dramatic, immediate failures, but after a few hits they cause equipment to fail prematurely. Get a good double conversion UPS and then check your grounding.
it could definitely be a drive issue as well. Awhile ago we had apretty bad batch of Western Digital IDE drives, only issues with those drives, none of our other drives, etc. Since then we've been using Seagates with minimal failures. Drives have moving parts and can fail, I wouldn't jump to conclussions.
Karl is correct. I have had bad hardware straight from the factory too. I have also had data centers that would quickly ruin equipment because of power and grounding problems. If you have performance problems in a variety of devices or simultaneous failures across multiple platforms I would check out the electrical environment. Regardless, a good UPS and a solid grounding system can't hurt.
Also, regarding your initial question, do you have reason to suspect that the data center temperature is a factor? Temperature is obviously very important.
egallant
Are the drives of the same brand?
What is the standing temp?
I've seen an entire batch of Maxtor IDE's come out bad just as gallant and KarlZimmer mentioned. It was a fault in the ball bearings from what we could figure out.
Heat will cause horrible things to happen to drives as well...Make sure if you are running raid arrays to do EVERYTHING possible to keep the machines and specifically the drives well ventilated.
Going as far as to keep those systems as close to the floor as possible is not a bad idea either.
I have found that SCSI drives are more reliable than ide drives. If you are running ide, you would be better off with the scsis. Do datacenters typically use IDE or SCSI?
Power and or heat are the likely culprits. In a small datacenter I used to manage we would have a chilled H20 supply problem on weekends and the datacenter temp would rise to the mid or high 70s Although the internal temps of the drives the remained in the "safe" zone, such an event would inevitably be followed by the failure of a couple of drives in the hotter zone of the drive arrays over the course of the following week.