VNXe 3300 monitoring snmp

Thank you !

I will download and try.

regards,

Tim

Rainer,

It is my understanding that even when using the UIMCLI - you are unable to get performance metrics like IOPS - out of the VNXe V1 platforms.  It is my understanding that the VNXe V1 platforms, like the 3100/3150/3300, are simply unable to provide performance metrics.

Also, using Unisphere Remote will not provide you performance statistics. It will only provide you the same minimal performance statistics you can glean from the VNXe V1 platform GUI - CPU Activity, Network Activity and SPA/SPA Volume Activity.

EMC released the VNXe V2 platform model 3200 in April 2014 which provides more performance information as shown here:

But if you are running a V1 VNXe platform, you are out of luck besides what is available in the GUI and what is shown in this blog:

http://henriwithani.wordpress.com/2011/12/01/hidden-vnxe-performance-statistics/

I cannot think of one network device, besides the VNXe V1 platforms, that I have managed in 10 years that didn't provide SNMP polling support.  A glaring omission in my opinion.

Amir

I would suggest to file a product enhancement request

Do you have any guides on how to use the uemcli to get the performance data off the system?

Hi Scott,

I've already been down this road so I thought I would spare you the suffering.  Unlike most network based platforms that have been produced since - oh I'd say the Cisco 2501 router : ) - the VNXe doesn't provide SNMP polling support.  This is, of course, the most common way to extract performance information from a platform.

So you will have to use a combination of linux commands (see CPU example) and UEMCLI commands (see Storage Utilization example) and ultimately if you want IOPS performance metrics - you will have no choice but to follow the superb recommendations of this blogger: henriwithani.wordpress.com:

LINUX COMMAND EXAMPLE

CPU

Given the lack of SNMP polling support, the Linux "top" command will be utilized, in batch mode, to acquire the combined CPU utilization of each Service Processor as follows:

1. ssh to “Management IP” of EMC VNXe Platform using the "service" credential as shown here:

ssh service@10.160.54.20 Welcome to EMC VNXe3300 12GB - (none) - NeoMain-2.4.2.21519-MAGNUM-RETAIL Password: Last login: Thu Jan  8 21:58:53 2015 from 172.19.39.71

2. Issue the CLI command top command in batch mode by using the -b option and specifying a specific number of iterations by using the -n option and dump to a file.

service@(none) spb:~> top -b -n 1 > ~/cpu.txt

3. Top output, by default, provides the average utilization of all CPUs as shown below:

service@(none) spb:~> ssh peer Last login: Wed Jan 28 17:13:38 2015 from peer service@(none) spa:~> top -b -n 1 > ~/cpu.txt service@(none) spa:~> cat ./cpu.txt top - 17:15:25 up 46 days, 20:37,  2 users,  load average: 67.78, 67.65, 67.46 Tasks: 267 total,  1 running, 266 sleeping,  0 stopped,  0 zombie Cpu(s):  4.3%us,  2.7%sy,  0.0%ni, 92.4%id,  0.0%wa,  0.0%hi,  0.5%si,  0.0%st Mem:  2138800k total,  2057556k used,    81244k free,  265700k buffers Swap:  1092412k total,      40k used,  1092372k free,  902996k cached 

4. Issue "ssh peer" to access other Service Processor and repeat steps 2 and 3 above:

 service@(none) spb:/var/log/sa> ssh peer Last login: Wed Oct  1 20:05:25 2014 from peer service@(none) spa:~>

NOTES:

• Create a graph for each Service Processor called "$SPX CPU Usage" - where X is either A (for spa) or B (for spb).
• For each graph, the max value should be 100% CPU Utilization
• For each graph,  take the %id value and subtract from 100 and plot
• Plot all remaining values
• Create a threshold of 90% CPU Utilization

UEMCLI EXAMPLE

VMware NFS Datastore Utilization

If the VNXe has been configured for VMware NFS Datastore storage, install the UEMCLI package and issue the following command to determine the VMware NFS Datastore Utilization:

uemcli -d X.X.X.X  -u Local/admin -p PASSWORD# /stor/prov/vmware/nfs show

Storage system address: X.X.X.X

Storage system port: 443

HTTPS connection

1:    ID                   = app_1

      Name                 = NFSXXX15K_SPA

      Description          =

      Health state         = OK (5)

      Server               = file_server_0

      Storage pool         = StoragePool00

      Size                 = 1408474546176 (1.3T)

      Size used            = 1117440180224 (1.0T)

      Cached               = no

      Protection size      = 0

      Protection size used = 0

      Local path           = /NFSXXX15K_SPA

      Export path          = /NFSXXX15K_SPA

NOTES:

• Create a graph for each NFS Share name called “$VMwareNFS_X Disk Space" - where X is the unique NFS Share name identified in the "Name" element
• Max value is equal to "byte" value in the "Size" field, converted to Gigabyte equivalent
• Create a threshold of 90% of Max value
• Poll and plot as the current VMware NFS Datastore Utilization, on a five minute interval, the "byte" value in "Size used" field, converted to Gigabyte equivalent

Henriwithani EXAMPLE

Hidden VNXe performance statistics | Henriwithani

Good luck and please rate helpful posts.

Amir Safayan

Hi all,

I have to extract VNXe performance metrics using UEMCLI commands.

For VNXe 1600, I am able to get paths for all performance metrics as it is a VNXe generation 2 device using commands /metrics/metric but for VNXe 3100 and VNXe3300 the command is showing Object not found error.

Please let me know if there are some other commands for last 2 mentioned devices to get performance data or the command is same and its some configuration issue or somethings needs to be enabled at storage device end.