Dell PowerEdge MX750c EMC PMem 200 Series Users Guide - Page 41

Delete namespaces, Management utility, Check PMem health status, Linux errata

Page 41 highlights

To write data into the device, run the following command: cd /mnt/pmemX echo "Hello World" >>test.txt Reset the system and the data should be persistent over power cycle. Delete namespaces Namespace can be deleted using NDCTL command: ndctl destroy-namespace is namespaceX.Y device name which can be retrieved using the following command: ndctl list -N To delete all the namespaces on the system, you may have to send the command multiple times NOTE: Reconfiguration of the goal or region from BIOS setup will automatically delete all the namespace on the system. Management utility Linux distributions use native tool ndctl to manage nonvolatile devices. For comprehensive list of commands and syntax, see open-source resource pmem.io. Check PMem health status Following NDCTL command shows the health status of each PMems installed on the system: ndctl list -DHi Health Information includes: Table 13. Health information Health fields Overall Health State DIMM Temperature Spare Percentage/Lifetime percentage User configured alarm temperature User configured alarm spares Shutdown state Comment Critical, Non-critical, ok In degree C In % Smart controller temperature threshold for alarm Smart spares threshold for alarm Dirty or Clean. Determines if data has been successfully cleaned to persistent domain during previous power cycle. Linux errata 1. If an uncorrectable error occurs in the early metadata region of the PMem, the system can get into an infinite loop of being unresponsive when booting into Linux. Workaround: Boot with "modprobe.blacklist=nd_pmem" on the kernel command line to stop the consumption of the error and enable boot progress. Afterwards, wait sufficient time for the Address Range Scrub feature to scrub the memory and identify the bad address location so it will be mapped out on future boots. 2. Linux boot fails to Emergency Mode when PMem is configured to automount during boot in fstab. Linux 41

  • 1
  • 2
  • 3
  • 4
  • 5
  • 6
  • 7
  • 8
  • 9
  • 10
  • 11
  • 12
  • 13
  • 14
  • 15
  • 16
  • 17
  • 18
  • 19
  • 20
  • 21
  • 22
  • 23
  • 24
  • 25
  • 26
  • 27
  • 28
  • 29
  • 30
  • 31
  • 32
  • 33
  • 34
  • 35
  • 36
  • 37
  • 38
  • 39
  • 40
  • 41
  • 42
  • 43
  • 44
  • 45
  • 46
  • 47

To write data into the device, run the following command:
cd /mnt/pmemX
echo “Hello World” >>test.txt
Reset the system and the data should be persistent over power cycle.
Delete namespaces
Namespace can be deleted using NDCTL command:
ndctl destroy-namespace <namespace>
<namespace> is namespaceX.Y device name which can be retrieved using the following command:
ndctl list -N
To delete all the namespaces on the system, you may have to send the command multiple times
NOTE:
Reconfiguration of the goal or region from BIOS setup will automatically delete all the namespace on the system.
Management utility
Linux distributions use native tool ndctl to manage nonvolatile devices.
For comprehensive list of commands and syntax, see open-source resource
pmem.io
.
Check PMem health status
Following NDCTL command shows the health status of each PMems installed on the system:
ndctl list -DHi
Health Information includes:
Table 13. Health information
Health fields
Comment
Overall Health State
Critical, Non-critical, ok
DIMM Temperature
In degree C
Spare Percentage/Lifetime percentage
In %
User configured alarm temperature
Smart controller temperature threshold for alarm
User configured alarm spares
Smart spares threshold for alarm
Shutdown state
Dirty or Clean. Determines if data has been successfully
cleaned to persistent domain during previous power cycle.
Linux errata
1.
If an uncorrectable error occurs in the early metadata region of the PMem, the system can get into an infinite loop of being
unresponsive when booting into Linux.
Workaround:
Boot with "
modprobe.blacklist=nd_pmem"
on the kernel command line to stop the consumption of the error
and enable boot progress. Afterwards, wait sufficient time for the Address Range Scrub feature to scrub the memory and
identify the bad address location so it will be mapped out on future boots.
2.
Linux boot fails to Emergency Mode when PMem is configured to automount during boot in fstab.
Linux
41