RAID related guidence required
Khurram Shahzad asked what happens on RAID 5 SSDs on a SAN if a commit write times out and the rollback write also fails. Art Kagel said it was unlikely but the database would be damaged, and strongly opposed RAID5. Neil Truby noted this holds for any RAID level; Art agreed but argued such failures are more common with RAID5 than RAID10.
Auto-generated by Claude from the posts below — may be imperfect; read the full thread.
We are planning to shift our database to RAID 5 on SSDs on VNX 5400 SAN. We are curious about the following scenario please guide us. If we are in the middle of a transaction, and the database needs to commit the data, and the write times-out because of RAID 5 failure, what happens ? and then if the rollback fails because of a write time-out, are we stuck with an inconsistent database state ?
Unlikely scenarios, but yes, you would be hosed. Have you read my RAID5 Rant? I don't care what the SaAN/NAS vendor said, RAID5 is NOT SAFE for your data. Periid, end of discussion. NO RAID5!!! NO RAID5!!! NO RAID5!!! NO RAID5!!! NO RAID5!!! NO RAID5!!! Art On Jun 18, 2015 5:09 AM, "KHURRAM SHAHZAD" <kshahzad02@i2cinc.com> wrote: > We are planning to shift our database to RAID 5 on SSDs on VNX 5400 SAN. We > are curious about the following scenario please guide us. > If we are in the middle of a transaction, and the database needs to commit > the > data, and the write times-out because of RAID 5 failure, > what happens ? and then if the rollback fails because of a write time-out, > are > we stuck with an inconsistent database state ? > > > > ******************************************************************************* > Forum Note: Use "Reply" to post a response in the discussion forum. > > --047d7bea35c063f9fd0518c80e5d
Your answer to the original post ("If the rollback fails because of a [RAID-5] write time-out, are we stuck with an inconsistent database state?") is correct of course, but is true of write failure at ANY RAID level. I do appreciate that your main message is that you believe that RAID-5 is more prone to such failures.
Neil: Yes, you are absolutely correct. Any completely failed write during a commit might result in data loss regardless of the RAID level. Indeed my contention is that such failures are more likely and indeed more common (though rare) under RAID5 than RAID10. Art Art S. Kagel, President and Principal Consultant ASK Database Management www.askdbmgt.com Blog: http://informix-myview.blogspot.com/ Disclaimer: Please keep in mind that my own opinions are my own opinions and do not reflect on the IIUG, nor any other organization with which I am associated either explicitly, implicitly, or by inference. Neither do those opinions reflect those of other individuals affiliated with any entity with which I am affiliated nor those of the entities themselves. On Fri, Jun 19, 2015 at 10:15 AM, NEIL TRUBY <neil.truby@ardenta.com> wrote: > Your answer to the original post ("If the rollback fails because of a > [RAID-5] > write time-out, are we stuck with an inconsistent database state?") is > correct > of course, but is true of write failure at ANY RAID level. > > I do appreciate that your main message is that you believe that RAID-5 is > more > prone to such failures. > > > > ******************************************************************************* > Forum Note: Use "Reply" to post a response in the discussion forum. > > --bcaec516235d2640ad0518e07fc8