Two SSD Failure on 2 Different Server(Single SSD) | Nutanix Community
Skip to main content
Solved

Two SSD Failure on 2 Different Server(Single SSD)

  • October 8, 2018
  • 3 replies
  • 1555 views

Hi Everyone, im just cruious about this scenario,

I have 3 server type 2310 with single ssd,
So if SSD FAIL on 2 different server at the same time what will happening?

Best answer by ddubuque

If it's a three node cluster and you lose two SSDs the short answer is the cluster will fail. When you loose one SSD, it will rebuild it's stored extents on the other two SSDs to maintain a resilient state. If you lost another SSD, there isn't enough to maintain the cluster resilience all storage I/O will fail because the CVM actually lives on the SSD. The hypervisor lives on the satadom or M.2 drive depending on your generation of hardware. You loose both SSDs and you're not only looking at data loss but complete Storage I/O communication cessation.
View original
Did this topic help you find an answer to your question?
This topic has been closed for comments

3 replies

  • Author
  • Voyager
  • 2 replies
  • October 22, 2018
Anyone can answer it?

ddubuque
Forum|alt.badge.img+10
  • Outrider
  • 63 replies
  • Answer
  • January 10, 2019
If it's a three node cluster and you lose two SSDs the short answer is the cluster will fail. When you loose one SSD, it will rebuild it's stored extents on the other two SSDs to maintain a resilient state. If you lost another SSD, there isn't enough to maintain the cluster resilience all storage I/O will fail because the CVM actually lives on the SSD. The hypervisor lives on the satadom or M.2 drive depending on your generation of hardware. You loose both SSDs and you're not only looking at data loss but complete Storage I/O communication cessation.

  • Author
  • Voyager
  • 2 replies
  • January 10, 2019
ddubuque wrote:
If it's a three node cluster and you lose two SSDs the short answer is the cluster will fail. When you loose one SSD, it will rebuild it's stored extents on the other two SSDs to maintain a resilient state. If you lost another SSD, there isn't enough to maintain the cluster resilience all storage I/O will fail because the CVM actually lives on the SSD. The hypervisor lives on the satadom or M.2 drive depending on your generation of hardware. You loose both SSDs and you're not only looking at data loss but complete Storage I/O communication cessation.

Yes You Right Sir,
Im Using Lenovo for the server and the problem is 2 SSD on 2 different Nodes failure.
Because Of that we lose our data then.

After discuss with Lenovo team finally we rebuild from the start using 2 ssd for every nodes, so SSD can Redudant each other for miimalism loss data.

Conclusion :
  • If you try bought 3 nodes server make sure every nodes had 2 ssd for minimalism Loss Data
  • Using Third Party Backup Program for keep your data safe.