I have a home server with many uses, but its main use is to hold all the storage I need. The hardware is pretty old: an Intel mobo with a Core 2 6600 and a NVidia 7600GS, to which there are 6 drives attached:
- 1x small Hitachi Deskstar 140GB system drive (1)
- 2x WD Caviar Green 2TB EARX model (last year's models)
- 1x WD Caviar Green 2TB EZRX model (current "advanced format" model)
- 1x WD Caviar Green 3TB EZRX model
- 1x Toshiba 3TB drive
And this is all driven by a rather old 500W PSU.
Lately I've been getting some hardware errors from the drives: both EZRX drives are spewing hardware access errors into the logs, and today one of the EARX drive has started active up as well (2). I ran SMART self-tests on all drives, and they all came up clean, though the SMART logs for the problematic drive lists the errors that I've seen - which is why I believe the problem is not the actual drive hardware.
One of the suggestion was to upgrade the PSU. I bought a Cooler Master GXII 550W PSU. Is that enough or do I need more power?
(1) it has some bad sectors but its doing well enough and it still has around 100 or so spare sectors.
(2) errors look like this in dmesg:
ata9.00: exception Emask 0x0 SAct 0x0 SErr 0x0 action 0x0
ata9.00: irq_stat 0x40000001
ata9.00: failed command: READ DMA EXT
ata9.00: cmd 25/00:08:00:08:00/00:00:20:00:00/e0 tag 0 dma 4096 in
res 61/04:08:00:08:00/00:00:20:00:00/e0 Emask 0x1 (device error)
ata9.00: status: { DRDY DF ERR }
ata9.00: error: { ABRT }
ata9.00: failed to enable AA (error_mask=0x1)
ata9.00: failed to enable AA (error_mask=0x1)
ata9.00: configured for UDMA/133 (device error ignored)