items: 0
     
     
HP MDL (Midline) and FATA Disks    
     
Advanced Replacements
Online Warranty Checker
Online RMA Request
   
Presales Support

We offer no-charge pre-sales support for locating the correct replacement disk drive for your SAN/Workstation/Server.

Call Us:
888-878-4167 x1

E-Mail Us: sales@istoragenetworks.com

Models

1TB FATA (EVA x400)
AG691A

1TB FATA (EVA x100):
AG883A

500GB FATA (EVA x100):
370790-B22

1TB SATA MDL (MSA, Proliant)
454146-B21

750GB SATA MDL
458930-B21

500GB SATA MDL
458928-B21

 


HP Midline Disk Drives

Large Capacity, Low-Cost Bulk Storage for HP Proliant and EVA.

HP MDL Disk Drives for Proliant and EVA

What are Midline and FATA Disk Drives

FATA and Midline drives are a tier below enterprise class 10K and 15K drives. Usually spinning at 7200 RPM, HP Midline / FATA disk drives use the same inexpensive and dense media found in workstation class SATA disk drives. On the EVA, HP has its OEM (Seagate) attach a Fibre Channel PCB (Printed Circuit Board) to a SATA HDA (head disc assembly), thus the acronym F-ATA. SAS MDL drives operate in the same fashion, only with a SAS PCB instead of FC. Other HP drives (such as those used in the MSA2000) use an interposer card to attach SATA drives to a dual port SAS controller. The benefit of these hybrid drives is dual-port (redundant) connectivity as well as SCSI error checking and correction which is greatly superior to the algorithms in the ATA standard.

Notice in Picture 1 how the HDA is identical between the FATA drive (left) and SATA drive (right.) Many lower-end HP MDL disks do not have an HP logo. The media is just as failure prone between both drives, BUT the SCSI (FC) error correction on the FATA drives makes data loss slightly less probable. The FATA drive also has two data channels on the SCA-40 Pin connector.

Notice in Picture 2 the INTERPOSER used to convert a single port (data channel) SATA disk to a dual port disk. This Interposer was used on HP P2000 / MSA2000 SANs. Even though it has a SCA-40 PIN connector, NO conversion to the FC Protocol is done. The Sierra-Logic chip is simply a multiplexer.

Picture 3 shows the native SCA-40 Pin FC connector on a FATA drive. SAS-NL drives will have a standard dual port SAS connector as well.

Picture 1: SATA vs. FATA Head Disc Assembly Picture 2: HP SATA Interposer Picture 3: FATA Connector
HP SATA vs FATA HP MSA2000 / P2000 Interposer Card HP Native FATA Connector

Avoid New Bulk and New Pulls

Midline and FATA disk drives use the same low-cost dense media found in desktop SATA drives. These drives are best suited for archival and non-random I/O. MDL drives have a short life span and tend to start growing defects after 2-3 years of use. As such, extra care should be avoided when buying any drives that are not FACTORY SEALED in HP boxes or refurbished by a qualified repair facility. Below are images from test software showing the true condition of a "new bulk" drive we have purchased:

ST31000340NS Failed SMART
  • This "new bulk" drive has 20 grown defects. New drives will have 0.
  • This "new bulk" drive has 27,806 power on hours.
  • This "new bulk" drive has uncorrectable errors. These errors mean data was written but cannot be retrieved. This drive should be RMAed ASAP or recycled as it is incapable of reliably storing data.

 

What to expect from an iStorage Networks Disk Drive

iStorage Networks verifies that every drive we ship is either new (as identified by the number of power on hours being less than 5) or being a qualified refurbished unit. All our drives will ship with the latest firmware revision and will never have any reallocated sectors (aka grown defects) or sectors pending reallocation. Grown defects are the number one indicator of a failing disk drive (see a great Google case study here.)

iStorage Networks will never ship a disk drive with any SMART parameter out of Seagate specification.

ST31000340NS SMART Pass
  • Most of our drives have under 100 power on hours (including refurbished units.)

  • A good drive (whether new or refurbished) should have 0 sectors pending reallocation.
  • ECC errors are normal. Unrecoverable errors should be 0.