Deferred Disk Service

Jump to solution

Hi,

Could some one throw some details on the following aspects regarding Deferred Disk Service?

1. Is there any advantage apart from the reduced number of CE visits, which in turn save on cost?

2. DDS is to ensure that the failed spares (??) would be replaced by CE in a bulk.

               2.1 Is there a fixed numeric value as to how many drives should fail before the CE replaces them all?

               2.2 Is it always true that the CE replaces the drives in bulk when there is just one spare left in the array?

               2.3 Is DDS going to take care of the disk technology, or is it just the covering spare count that matters?         

3. Would it not be a risk if a particular disk with a specified part number is not available readily, whilst the CE needs to replace the disks in bulk? Is there some mechanism put in place to ensure that this risk is mitigated?

4. Is there any difference between DMX and VMAX in the way the DDS operates? Will VMAX also undergo standard sparing with DDS enabled under any scenario?

Thanks,

Sreehari

Labels (1)
0 Kudos
2 Replies
PaulCork
3 Argentium

Re: Deferred Disk Service

Jump to solution
0 Kudos
Highlighted

Re: Deferred Disk Service

Jump to solution
  1. Is there a fixed numeric value as to how many drives should fail before the CE replaces them all?
    1. There are 2 settings:  0 or 1.  When there is either 0 or 1 Spare left, DDS will send a notification of the Drives to the Field for replacement
  2. Is it always true that the CE replaces the drives in bulk when there is just one spare left in the array?
    1. If Deferred Service is dependent on the Spare Pool.  If there are EFD, 10k and 15k Technology in the Array and each type has 3 covering spares, if the 15k pool drops to the setting, 0 or 1, then an SR will be sent for replacement of ALL drives so it is dependent on the Spare Pool and not the total number of failed drives
  3. Is DDS going to take care of the disk technology, or is it just the covering spare count that matters?
    1. Spare Pool covers disk technology
  4. Would it not be a risk if a particular disk with a specified part number is not available readily, whilst the CE needs to replace the disks in bulk? Is there some mechanism put in place to ensure that this risk is mitigated?
    1. The only time there would be a risk is if the Drive is a standalone, meaning it is not part of a RAID Group.  When CE’s come onsite to replace failed drives, part of the replacement script checks to ensure that there is no possibility of Data Loss.  If there is a potential for Data Loss, the script will notify the CE and is some instances will give the CE the option of acknowledging data loss by typing DATA LOSS in the script.
    2. The Mechanism that is used is called RAID, Redundant Array of Independent Disks.  EMC uses RAID 5 and RAID 6, 3+1 and 5+2 respectively.  For RAID 5 that are 4 drives in the RAID Group and the data is copied between 3 of the drives with a complete copy on 1.  We can have 1 failed drive in a RAID 5 setup without any problems. (a failed drive is a drive that has failed but does NOT have a Spare invoked)  RAID 6 is the same as 5, however there are 7 drives with data spread across 5 and 2 copies.  With RAID 6 we can lose 2 drives without any problem.
  5. Is there any difference between DMX and VMAX in the way the DDS operates? Will VMAX also undergo standard sparing with DDS enabled under any scenario?
    1. Yes, with DMX we don’t use DDS until 5773.176 and newer.  With VMax all Codes use DDS.  ALL VMax’s will use Sparing regardless of DDS being enabled.  The only difference between DDS enabled and disabled is that either a single drive is replaced as it fails or multiple drives are replaced at once.
0 Kudos