we come from dm_cache and have set up our first node (hosting KVM machines with SolusVM) with EnhanceIO after successful tests and benchmarks which showed much better performance. The node has been stable for two weeks and then crashed twice within four days. It looked like this:
The setup uses a Software-RAID-1 with two SSDs as cache in writeback mode and a Hardware-RAID-10 as actual storage.
Are there any suggestions? The first thing we will test is going back to kernel 3.14.39 this night, but to be honest, I don't think that this issue is directly related to the newer kernel version we used to use.
Might it be an idea to change the policy (LRU) or switch the mode (WriteBack)?
Hi,
we come from dm_cache and have set up our first node (hosting KVM machines with SolusVM) with EnhanceIO after successful tests and benchmarks which showed much better performance. The node has been stable for two weeks and then crashed twice within four days. It looked like this:
The setup uses a Software-RAID-1 with two SSDs as cache in writeback mode and a Hardware-RAID-10 as actual storage.
Are there any suggestions? The first thing we will test is going back to kernel 3.14.39 this night, but to be honest, I don't think that this issue is directly related to the newer kernel version we used to use.
Might it be an idea to change the policy (LRU) or switch the mode (WriteBack)?
Thanks in advance!
Best regards, Tim