SNMP monitoring Nimble storage

[petr-macek]

Hello,

NIMBLE SNMP mib contains information about volume size and usage. Volume size isn't problem. It is here:

.1.3.6.1.4.1.37447.1.2.1.4.0 = Gauge32: 5242880  -- first volume 5TB
.1.3.6.1.4.1.37447.1.2.1.4.1 = Gauge32: 512000 -- second volume 500GB

But volume usage? There are more OIDs:

volDiskVolBytesUsedLow .1.3.6.1.4.1.37447.1.2.1.52
volDiskVolBytesUsedHigh .1.3.6.1.4.1.37447.1.2.1.53
volDiskSnapBytesUsedLow .1.3.6.1.4.1.37447.1.2.1.54
volDiskSnapBytesUsedHigh .1.3.6.1.4.1.37447.1.2.1.55
volUsageLow .1.3.6.1.4.1.37447.1.2.1.6
volUsageHigh .1.3.6.1.4.1.37447.1.2.1.7

 I can't find a volume usage value that matches the GUI data

[buzzsubash]

Hello petr-macek,

Can you please let me know the version of Nimble OS running on the array ? Also, which tool is being used for monitoring ?

[petr-macek]

Hello,

it is NIMBLE HF20H. Version from snmp .1.3.6.1.2.1.1.1.0 = STRING: Nimble Storage Grp-nimble1 running software version 5.2.1.1000-1017822-opt

I'm Cacti monitoring tool developer and I want to create template for Nimble device. Now I'm trying only SNMP walk.  Maybe I can't evaluate values properly, MIB is not very detailed.

Current state:

From SNMP I got these values. It is global statistics. These numbers do not correspond with screenshot (volume usage).

.1.3.6.1.4.1.37447.1.3.12.0 = Gauge32: 329219025  -- Total number of bytes used on disk for volumes (low order bytes). 
.1.3.6.1.4.1.37447.1.3.13.0 = Gauge32: 368 -- Total number of bytes used on disk for volumes (high order bytes)
.1.3.6.1.4.1.37447.1.3.14.0 = Gauge32: 2962999749 -- Total number of bytes used on disk for snapshots (low order bytes).
.1.3.6.1.4.1.37447.1.3.15.0 = Gauge32: 701 -- Total number of bytes used on disk for snapshots (high order bytes)

To be honest, I don;t understand Low/Hi values. For size I need only one number normally

Volume specific data:

VolSizeLow:

.1.3.6.1.4.1.37447.1.2.1.4.0 = Gauge32: 5242880 ---  value/1024/1024 = 5T - correct
.1.3.6.1.4.1.37447.1.2.1.4.1 = Gauge32: 512000 -- value/1024 = 500G - correct

VolSizeHi:.

1.3.6.1.4.1.37447.1.2.1.5.0 = Gauge32: 0
.1.3.6.1.4.1.37447.1.2.1.5.1 = Gauge32: 0

I don' understand to other numbers:

VolUsageLow (Usage in MB (low order bytes) )

.1.3.6.1.4.1.37447.1.2.1.6.0 = Gauge32: 1506294  -- /1024/1024 = 1,4TB, but nimble says 3TB
.1.3.6.1.4.1.37447.1.2.1.6.1 = Gauge32: 1347 -- /1024 = 1.3G, but nimble says 3.5GB

These numbers don't make sense to me too:

volDiskVolBytesUsedLow Unsigned32 Total number of bytes used on disk for volumes - low order bytes.

volDiskVolBytesUsedHigh Unsigned32 Total number of bytes used on disk for volumes - high order bytes.

volDiskSnapBytesUsedLow Unsigned32 Total number of bytes used on disk for snapshots - low order bytes.

volDiskSnapBytesUsedHigh Unsigned32 Total number of bytes used on disk for snapshots - high order bytes.

[buzzsubash · Accepted solution]

Thank you for letting know.

Can you please perform the following steps to convert the high order bytes and low order bytes into a usable number.  Note that performing the steps in reverse order should allow you to take the volume usage (in MegaBytes) and predict the SNMP values for the "UsedLow" and "UsedHigh" OIDs.

Conversion steps, SNMP values to readable volume usage:

1. Convert the lower order number to binary (make sure it's padded to 32 bits)                                 
2. Convert the high order number to binary
3. Append the low order bits after the high order bits
4. Convert the binary to decimal to get the number of bytes
5. divide the bytes by (1000*1024) to get megabyte .

If the above steps does not solve the issue, please do DM the serial number of the array along with your Email ID. I shall create a case and take ownership of it to continue working with you.

[support_s]

Hello,

Let us know if you were able to resolve the issue.

If you have no further query, and you are satisfied with the answer then kindly mark the topic as Solved so that it is helpful for all community members.

Please click on "Thumbs Up/Kudo" icon to give a "Kudo".

Thank you for being a HPE valuable community member.

[patiljuwel]

Thank you for the excellent Advise, @buzzsubash . 

I would like to know - how can I use Grafana to showcase these hardware parameters in a dashboard ?

I have prometheus data source pre-configured and we use it for monitoring the storage specs on-prem. 

Thank you.

[Mr_Techie]

@patiljuwel 

Verify that hardware-related metrics are being exported to Prometheus. Common exporters include:

Node Exporter: For general hardware metrics (CPU, memory, disk usage, network).
IPMI Exporter: For server-specific metrics like temperatures, fan speeds, and power status.
SNMP Exporter: For devices supporting SNMP (like storage systems or switches).
Custom Exporters: If specific hardware uses proprietary APIs.

Use Prometheus's UI or queries to confirm which metrics are available, then set-up a Grafana Dashboard

Customize the panels

Use different visualization types:
Graph: For trends like CPU and memory usage.
Gauge/Stat: For point-in-time values like temperature or fan speed.
Table: For showing multiple metrics side-by-side.
Add thresholds (e.g., turn red if temperature exceeds safe limits).

Configure alerts in Grafana based on metrics,

Use templating variables to filter metrics dynamically by server, hardware type, or location.
Create repeatable panels if monitoring multiple devices with similar metrics.

If your storage system is already monitored via Prometheus, include those metrics in the same dashboard to have a unified view of hardware and storage performance.

[AlexeiR267]

We play with it today.

Volume usages and volume sizes are in MB (not in bytes), and values match GUI. Vol usage is LOW * 1024*1024  and Vol size is LOW * 1024*1024

used Bytes is total usage in BYTES (so usage-bytes = (HIGH * 2^32 + LOW)) and match sum(VOLUMES / COMPRESSION) (but do not match deduplication outcome).

No one MIB shows FREE space which is critical (and which do not match these numbers). 

So NIMBLE MIB is total mess, I managed to create zabbix template for it (can post it) but it's not of a great use as it miss the most important number (total free space in the group). Anyway, it works, more or less.

[petr-macek]

I wanted it for the Cacti template. Finally I created the template, with similar results and findings as you