• Visitors can check out the Forum FAQ by clicking this link. You have to register before you can post: click the REGISTER link above to proceed. To start viewing messages, select the forum that you want to visit from the selection below. View our Forum Privacy Policy.
  • Want to receive the latest contracting news and advice straight to your inbox? Sign up to the ContractorUK newsletter here. Every sign up will also be entered into a draw to WIN £100 Amazon vouchers!

Reply to: HP 3PAR and SSD

Collapse

You are not logged in or you do not have permission to access this page. This could be due to one of several reasons:

  • You are not logged in. If you are already registered, fill in the form below to log in, or follow the "Sign Up" link to register a new account.
  • You may not have sufficient privileges to access this page. Are you trying to edit someone else's post, access administrative features or some other privileged system?
  • If you are trying to post, the administrator may have disabled your account, or it may be awaiting activation.

Previously on "HP 3PAR and SSD"

Collapse

  • WilsonPiper72
    replied
    Originally posted by sal View Post
    How big is the array? The 3PAR inForm OS works with chunklets, not disks so essentially your "RAID" spans the whole array, not just 10-12 disks which is the usual number of disks you will see in a RAID6 virtual disk. Meaning that the performance overhead is minimal on large arrays.

    RAID6 is the defacto king of storage when you are after maximising capacity, with RAID10 used for performance volumes.

    RAID5 is generally not recommended to production load. The problem with it is that with high capacity disk these days it might take days to rebuild the array after failure, leaving you without redundancy, exactly at a time of high stress during the rebuild. Not worth the risk for the expense of "losing" one more disk capacity with RAID6

    I have no idea how well the 3PAR handles SSD, last time I worked with one was back in 2013, when they were still relatively new on the market and SSDs weren't cheap enough for the average enterprise.
    Thank you...

    Leave a comment:


  • sal
    replied
    How big is the array? The 3PAR inForm OS works with chunklets, not disks so essentially your "RAID" spans the whole array, not just 10-12 disks which is the usual number of disks you will see in a RAID6 virtual disk. Meaning that the performance overhead is minimal on large arrays.

    RAID6 is the defacto king of storage when you are after maximising capacity, with RAID10 used for performance volumes.

    RAID5 is generally not recommended to production load. The problem with it is that with high capacity disk these days it might take days to rebuild the array after failure, leaving you without redundancy, exactly at a time of high stress during the rebuild. Not worth the risk for the expense of "losing" one more disk capacity with RAID6

    I have no idea how well the 3PAR handles SSD, last time I worked with one was back in 2013, when they were still relatively new on the market and SSDs weren't cheap enough for the average enterprise.

    Leave a comment:


  • WilsonPiper72
    replied
    Originally posted by malvolio View Post
    Ah, the Magic SAN consultancy. Complete black box, especially if you work with EMC (I once migrated several dozen Tb for them from one datacentre to another with no common infrastructure between them. And nobody had maintained the drive maps. That was fun...). Although, to be fair, they do adhere to the principles I outlined above, and properly built SAN arrays use RAID and real and virtual drive architectures in various combinations. At least you now know what questions to ask...

    Leave a comment:


  • malvolio
    replied
    Originally posted by WilsonPiper72 View Post
    Thanks malvolio for your time and detailed explanation ..
    As I mentioned in my opening statement I am not a storage specialists and my core skills are Oracle database(s)..

    usually in the past, I have in a association with ( team of 3 - Oracle DBA/ Unix Admin & Storage Admin) configured storage in a such way mainly RAID 10 so that it will be for both resilience and disaster tolerance. I know it is not cost effective but from business point of view one thing less to worry about.

    I will take you explanation on board and think more about because SAN Storage admin guys are from 3rd party and nothing is visible to us from our end.

    Cheers.
    W.
    Ah, the Magic SAN consultancy. Complete black box, especially if you work with EMC (I once migrated several dozen Tb for them from one datacentre to another with no common infrastructure between them. And nobody had maintained the drive maps. That was fun...). Although, to be fair, they do adhere to the principles I outlined above, and properly built SAN arrays use RAID and real and virtual drive architectures in various combinations. At least you now know what questions to ask...

    Leave a comment:


  • WilsonPiper72
    replied
    Originally posted by malvolio View Post
    RAID write penalty is about 10% depending on your write cache capability. It also depends on the hardware and configuration in use, but fast drives such as SSDs for new work and HDDs for long term storage is the optimum, with logging on to a separate stack so you can easily add capacity where growth is highest.

    RAID5 is not disaster tolerant, and does not constitute a business continuity solution. RAID6 merely extends the point at which a failure causes a problem.

    I'm a bit out of date, but always advocated a RAID5/6 structure at the virtual level supported on RAID 0/1 mirrored hardware for both resilience and disaster tolerance (I came back from holiday once to find a head-crashed systems disk and nobody had noticed...). That does optimise read speeds, but if you want RAID then you have to accept that write performance penalty.
    Thanks malvolio for your time and detailed explanation ..
    As I mentioned in my opening statement I am not a storage specialists and my core skills are Oracle database(s)..

    usually in the past, I have in a association with ( team of 3 - Oracle DBA/ Unix Admin & Storage Admin) configured storage in a such way mainly RAID 10 so that it will be for both resilience and disaster tolerance. I know it is not cost effective but from business point of view one thing less to worry about.

    I will take you explanation on board and think more about because SAN Storage admin guys are from 3rd party and nothing is visible to us from our end.

    Cheers.
    W.

    Leave a comment:


  • malvolio
    replied
    Originally posted by WilsonPiper72 View Post
    Yes I Have! and I wanted industry feedback and NOT Smart answers!
    RAID write penalty is about 10% depending on your write cache capability. It also depends on the hardware and configuration in use, but fast drives such as SSDs for new work and HDDs for long term storage is the optimum, with logging on to a separate stack so you can easily add capacity where growth is highest.

    RAID5 is not disaster tolerant, and does not constitute a business continuity solution. RAID6 merely extends the point at which a failure causes a problem.

    I'm a bit out of date, but always advocated a RAID5/6 structure at the virtual level supported on RAID 0/1 mirrored hardware for both resilience and disaster tolerance (I came back from holiday once to find a head-crashed systems disk and nobody had noticed...). That does optimise read speeds, but if you want RAID then you have to accept that write performance penalty.

    Leave a comment:


  • WilsonPiper72
    replied
    Originally posted by BR14 View Post
    R.T.F.M.
    Yes I Have! and I wanted industry feedback and NOT Smart answers!

    Leave a comment:


  • BR14
    replied
    R.T.F.M.

    Leave a comment:


  • WilsonPiper72
    started a topic HP 3PAR and SSD

    HP 3PAR and SSD

    Hello All,

    I am novice in Storage so please be gentle. I look after 50TB Oracle DW database and
    Information I have is 3rd party is offering RAID6 SSD for better performance.

    Has any one configured FRA onto RAID6 SSD , our current FRA size is 10TB.
    and daily ~ 1TB of archive logs get generated.

    RAID6 will have write penalty and most importantly why HP is promoting RAID6 ... does SSD failure rate is high? or 3PAR HP firmware doesn't support those SSDs gracefully?

    Your valuable inputs will be much appreciated..

    Cheers
    Bill

Working...
X