Score:-1

8 x 15k SAS hdd disk with raid 0 performance vs single SSD disk for database

jp flag

Which one has a better performance for database application (e.g. postgres) if we just consider read/write operations speed and not fault tolerance

  • 8 x 15k SAS hdd disk with raid 0
  • single SSD disk
Mikael H avatar
br flag
The question is pointless: Nobody in their right mind would run a production database off a RAID0, and nobody in their right mind would run a production database off a single SSD. But to entertain your thought experiment: Hard drives count their 4k IOPS in the low hundreds. SSDs count their 4k IOPS in the tens of thousands. Adjust for bus speeds and database page sizes and do your math.
Score:2
br flag

You've given us almost zero information to work from here sorry, but let's make some assumptions and see where that gets us.

15K SAS disks are usually good for about 180-200MBps reading and writing sequential data, for random they're terrible and the SSD will win but let's stick with sequential. So 8 x 200MBps is 1.6GBps, which shouldn't flood a dual-12Gbps-SAS-bus HBA, nor will it flood the likely PCIe 3.0 x 8/16 lane bus it's connected to.

A single 12Gbps SSD is likely to be able to sustain that 12Gbps link, so think in the region of 1.4GBps of sequential performance.

So purely in terms of sequential access they're about the same, roughly anyway. But in terms of random operation the SSD will destroy the HDD's, by at least an order of magnitude.

Of course if you go with NVMe then the SSD wins both tests easily.

Ali Eb avatar
jp flag
database operations (specially write) is more sequential or random access? @Chopper3
br flag
That entirely depends on what your DB is doing - it could be 100% sequential, 100% random or more likely a bit of both, we just don't know, you could do with knowing.
br flag
@Zac67 I specifically said the PCIe wouldn't be a blocker
mangohost

Post an answer

Most people don’t grasp that asking a lot of questions unlocks learning and improves interpersonal bonding. In Alison’s studies, for example, though people could accurately recall how many questions had been asked in their conversations, they didn’t intuit the link between questions and liking. Across four studies, in which participants were engaged in conversations themselves or read transcripts of others’ conversations, people tended not to realize that question asking would influence—or had influenced—the level of amity between the conversationalists.