Some "metrics" being reported as dimensions

[M-Bates]

A number of data types that are the same as Gauge or Counter are currently being extracted as dimensions rather than metrics.
The data types are defined in the IETF textual conventions files, HCNUM-TC

• CounterBasedGauge64
• ZeroBasedCounter64

[ajasnosz]

The values you mention as CounterBasedGauge64 are jnxJsNodeSessionCreationPerSecond and jnxJsNodeSessCreationPerSecIPv4 in this mib, I don't see any ZeroBasedCounter64, are there any of this type? Were they presented as metrics before? Could you show how they look like without using sc4snmp?

[M-Bates]

The values for the 3 CounterBasedGuauge64 from the Juniper MIB has always been interpreted as a dim rather than metric

• jnxJsNodeSessCreationPerSecIPv4: 13
• jnxJsNodeSessCreationPerSecIPv6: 0
• jnxJsNodeSessionCreationPerSecond: 13

snmpget -v 2c -c <> -m +JUNIPER-SRX5000-SPU-MONITORING-MIB <>.1.3.6.1.4.1.2636.3.39.1.12.1.4.1.5.0
JUNIPER-SRX5000-SPU-MONITORING-MIB::jnxJsNodeSessionCreationPerSecond.0 = Counter64: 3

Unfortunately, these are the only 3 CounterBasedGuage metrics being collected across all device types.

[ajasnosz]

If they were always interpreted as dim then it is what we expect, as this depends on mib and pysnmp library. We don't have any exceptions that we apply when translating, it's all automatically translated.

[M-Bates]

I cannot tell you whether they were always wrong.
What I can say is that they are wrong now, and need to be correctly sent as metric points not dimensions. This is also a function of this library, not pysnmp.

As requested, please fix the issue with the 2 data types being sent as dimensions rather than metric values.
This is a breaking issue as there is no way to consistently graph the values of dimensions over time for large volumes of data.

[ajasnosz]

I've prepared the image for sc4snmp with the code that might fix this. We don't have the devices with types you mentioned, so you can test the changes by adding below lines to the values.yaml and running the upgrade command. Let us know if this solution is changing the CounterBasedGuauge64 and ZeroBasedCounter64 to metrics.

image:
  # The registry and name of the SC4SNMP image to pull
  repository: ajasnosz558/sc4snmp-dev-hostname
  # The policy that specifies when the user wants the SC4SNMP images to be pulled
  pullPolicy: Always
  # Overrides the image tag whose default is the chart appVersion.
  tag: "new-metrics"

[ajasnosz]

@M-Bates did you have time to test above image?

[M-Bates]

Not as yet. The customer has been in a change freeze. I should the ability to test it around mid-Jan.

[ajasnosz]

@M-Bates was the image tested and is working fine?

[M-Bates]

Hi, I no longer have access to be able to test this. As the changes required are part of standard MIBS, not just Juniper specific, the changes should be testable and verifiable by the product owner. Failure of the product to correctly represent standard MIB data types is a fundamental failure in the product. Michael Bates Principal Consultant Mobile +61 411 440 095 · Phone (03) 9880 5528 Level 35, 80 Collins Street, Melbourne, VIC 3000, Australia · tesserent.com<https://tesserent.com/> [Tesserent] Follow Us [Linkedin] <https://www.linkedin.com/company/tesserent/> [Youtube] <https://www.youtube.com/channel/UC5SpXfc8EGshynZ3QSZYz_w>

…

________________________________ From: ajasnosz ***@***.***> Sent: Tuesday, 14 May 2024 18:39 To: splunk/splunk-connect-for-snmp ***@***.***> Cc: Michael Bates ***@***.***>; Mention ***@***.***> Subject: Re: [splunk/splunk-connect-for-snmp] Some "metrics" being reported as dimensions (Issue #918) @M-Bates was the image tested and is working fine? — Reply to this email directly, view it on GitHub, or unsubscribe. You are receiving this because you were mentioned. Message ID: <splunk/splunk-connect-for-snmp/issues/918/2109608785@ github. com> @M-Bates<https://urldefense.com/v3/__https://github.com/M-Bates__;!!CqhDG3qmu8NhxS99Ag!1SIrUDt56PLt4v_KoqME83ZVevtFoePRLOEaKBe-jmauXCFBTNp3NGaieHee76RaFGjzK0fVeprj2PPxsoj6V2OyC3EP$> was the image tested and is working fine? — Reply to this email directly, view it on GitHub<https://urldefense.com/v3/__https://github.com/splunk/splunk-connect-for-snmp/issues/918*issuecomment-2109608785__;Iw!!CqhDG3qmu8NhxS99Ag!1SIrUDt56PLt4v_KoqME83ZVevtFoePRLOEaKBe-jmauXCFBTNp3NGaieHee76RaFGjzK0fVeprj2PPxsoj6V-1FnTaq$>, or unsubscribe<https://urldefense.com/v3/__https://github.com/notifications/unsubscribe-auth/AU3U464FVRBUULNUQ5RSE3DZCHEURAVCNFSM6AAAAAA7O7XLUGVHI2DSMVQWIX3LMV43OSLTON2WKQ3PNVWWK3TUHMZDCMBZGYYDQNZYGU__;!!CqhDG3qmu8NhxS99Ag!1SIrUDt56PLt4v_KoqME83ZVevtFoePRLOEaKBe-jmauXCFBTNp3NGaieHee76RaFGjzK0fVeprj2PPxsoj6VxSimLTk$>. You are receiving this because you were mentioned.Message ID: ***@***.***>