Hi,
I know that in this forum people stress that ceph needs enterprise nvme disks, and infact I am using enterprise nvme with plp and so on.
But now I read this smart data from two clusters:
SMART/Health Information (NVMe Log 0x02)
Critical Warning: 0x00
Temperature: 38 Celsius
Available Spare: 100%
Available Spare Threshold: 10%
Percentage Used: 2%
Data Units Read: 178,846,698 [91.5 TB]
Data Units Written: 311,177,655 [159 TB]
Host Read Commands: 2,240,990,410
Host Write Commands: 14,861,274,116
Controller Busy Time: 3,052
Power Cycles: 14
Power On Hours: 10,818
Unsafe Shutdowns: 2
Media and Data Integrity Errors: 0
Error Information Log Entries: 0
Warning Comp. Temperature Time: 0
Critical Comp. Temperature Time: 0
Temperature Sensor 1: 38 Celsius
Temperature Sensor 2: 47 Celsius
SMART/Health Information (NVMe Log 0x02)
Critical Warning: 0x00
Temperature: 63 Celsius
Available Spare: 100%
Available Spare Threshold: 20%
Percentage Used: 1%
Data Units Read: 82,112,629 [42.0 TB]
Data Units Written: 124,809,695 [63.9 TB]
Host Read Commands: 662,799,183
Host Write Commands: 3,601,449,973
Controller Busy Time: 3,394
Power Cycles: 43
Power On Hours: 13,953
Unsafe Shutdowns: 38
Media and Data Integrity Errors: 0
Error Information Log Entries: 2
Warning Comp. Temperature Time: 1222
Critical Comp. Temperature Time: 4
Temperature Sensor 1: 57 Celsius
Temperature Sensor 2: 107 Celsius
Considering that these are light clusters with only ERP or other "mostly read" vms it seems to me that 159 written TB in one year are too much and are all made by Ceph.
I do not like very much the idea that I already have a 2% wear in an year of enterprise nvmes.
It seems that "enterprise nvme" is a requirements because ceph eats disks for breakfast.
Mario
I know that in this forum people stress that ceph needs enterprise nvme disks, and infact I am using enterprise nvme with plp and so on.
But now I read this smart data from two clusters:
SMART/Health Information (NVMe Log 0x02)
Critical Warning: 0x00
Temperature: 38 Celsius
Available Spare: 100%
Available Spare Threshold: 10%
Percentage Used: 2%
Data Units Read: 178,846,698 [91.5 TB]
Data Units Written: 311,177,655 [159 TB]
Host Read Commands: 2,240,990,410
Host Write Commands: 14,861,274,116
Controller Busy Time: 3,052
Power Cycles: 14
Power On Hours: 10,818
Unsafe Shutdowns: 2
Media and Data Integrity Errors: 0
Error Information Log Entries: 0
Warning Comp. Temperature Time: 0
Critical Comp. Temperature Time: 0
Temperature Sensor 1: 38 Celsius
Temperature Sensor 2: 47 Celsius
SMART/Health Information (NVMe Log 0x02)
Critical Warning: 0x00
Temperature: 63 Celsius
Available Spare: 100%
Available Spare Threshold: 20%
Percentage Used: 1%
Data Units Read: 82,112,629 [42.0 TB]
Data Units Written: 124,809,695 [63.9 TB]
Host Read Commands: 662,799,183
Host Write Commands: 3,601,449,973
Controller Busy Time: 3,394
Power Cycles: 43
Power On Hours: 13,953
Unsafe Shutdowns: 38
Media and Data Integrity Errors: 0
Error Information Log Entries: 2
Warning Comp. Temperature Time: 1222
Critical Comp. Temperature Time: 4
Temperature Sensor 1: 57 Celsius
Temperature Sensor 2: 107 Celsius
Considering that these are light clusters with only ERP or other "mostly read" vms it seems to me that 159 written TB in one year are too much and are all made by Ceph.
I do not like very much the idea that I already have a 2% wear in an year of enterprise nvmes.
It seems that "enterprise nvme" is a requirements because ceph eats disks for breakfast.
Mario