fix error

[insobi]

the 'None' value returned from regex.match() wasn't handled as expected in those if condition, so errors happened. As a workaround, the type of return value from regex.match() could convert as a string.

[RavuAlHemio]

Greetings!

I'm not entirely sure that this is the correct solution. There should be some differentiation between the null value (None in Python, null in JSON and YAML) and a string containing such a word ("None" or "null") -- otherwise somebody could name an ACI object "null" and this would cause problems with this exporter -- and your fix would remove this distinction.

Can you provide a configuration snippet with which you are having trouble?

[insobi]

Hi,
I posted below error log which happened from previous commit. Please refer to it.

DEBUG:prometheus_aci_exporter:processing instance of class 'eqptIngrBytes5min' with DN 'topology/pod-1/node-104/sys/ch/supslot-1/sup/cpuport-1/CDeqptIngrBytes5min'
DEBUG:prometheus_aci_exporter:indexing: index -1, suffix ''
DEBUG:prometheus_aci_exporter:processing label definition {'property_name': 'dn', 'regex': '^topology/pod-(?P<pod>[1-9][0-9]*)/node-(?P<node>[1-9][0-9]*)/sys/(?P<interfaceType>[a-z]+)-\\[(?P<interface>[^\\]]+)\\]/'} for attributes {'childAction': '', 'cnt': '14', 'dn': 'topology/pod-1/node-104/sys/ch/supslot-1/sup/cpuport-1/CDeqptIngrBytes5min', 'floodAvg': '0', 'floodBase': '0', 'floodCum': '0', 'floodLast': '0', 'floodMax': '0', 'floodMin': '0', 'floodPer': '0', 'floodRate': '0.000000', 'floodSpct': '0', 'floodThr': '', 'floodTr': '0', 'floodTrBase': '0', 'lastCollOffset': '140', 'modTs': 'never', 'multicastAvg': '0', 'multicastBase': '0', 'multicastCum': '0', 'multicastLast': '0', 'multicastMax': '0', 'multicastMin': '0', 'multicastPer': '0', 'multicastRate': '0.000000', 'multicastRateAvg': '0.000000', 'multicastRateLast': '0.000000', 'multicastRateMax': '0.000000', 'multicastRateMin': '0.000000', 'multicastRateSpct': '0', 'multicastRateThr': '', 'multicastRateTr': '0.000000', 'multicastRateTrBase': '0.000000', 'multicastRateTtl': '0.000000', 'multicastSpct': '0', 'multicastThr': '', 'multicastTr': '0', 'multicastTrBase': '0', 'repIntvEnd': '2021-07-29T21:12:16.354+09:00', 'repIntvStart': '2021-07-29T21:09:56.351+09:00', 'status': '', 'unicastAvg': '0', 'unicastBase': '0', 'unicastCum': '0', 'unicastLast': '0', 'unicastMax': '0', 'unicastMin': '0', 'unicastPer': '0', 'unicastRate': '0.000000', 'unicastRateAvg': '0.000000', 'unicastRateLast': '0.000000', 'unicastRateMax': '0.000000', 'unicastRateMin': '0.000000', 'unicastRateSpct': '0', 'unicastRateThr': '', 'unicastRateTr': '0.000000', 'unicastRateTrBase': '0.000000', 'unicastRateTtl': '0.000000', 'unicastSpct': '0', 'unicastThr': '', 'unicastTr': '0', 'unicastTrBase': '0'}
DEBUG:prometheus_aci_exporter:updated labels are 'topology/pod-1/node-104/sys/ch/supslot-1/sup/cpuport-1/CDeqptIngrBytes5min'
----------------------------------------
Exception happened during processing of request from ('127.0.0.1', 58520)
Traceback (most recent call last):
  File "prometheus_aci_exporter.py", line 354, in collect_fabric
    labels.update(updated_labels)
ValueError: dictionary update sequence element #0 has length 1; 2 is required

During handling of the above exception, another exception occurred:

Traceback (most recent call last):
  File "/usr/lib/python3.8/socketserver.py", line 683, in process_request_thread
    self.finish_request(request, client_address)
  File "/usr/lib/python3.8/socketserver.py", line 360, in finish_request
    self.RequestHandlerClass(request, client_address, self)
  File "/usr/lib/python3.8/socketserver.py", line 747, in __init__
    self.handle()
  File "/usr/lib/python3.8/http/server.py", line 427, in handle
    self.handle_one_request()
  File "/usr/lib/python3.8/http/server.py", line 415, in handle_one_request
    method()
  File "prometheus_aci_exporter.py", line 551, in do_GET
    generated_lines = [
  File "prometheus_aci_exporter.py", line 551, in <listcomp>
    generated_lines = [
  File "prometheus_aci_exporter.py", line 249, in collect
    yield from self.collect_fabric(fabric_name, fabric, controller, common_queries)
  File "prometheus_aci_exporter.py", line 356, in collect_fabric
    raise ValueError(f"failed to update labels in query {query_name!r} with {updated_labels!r}: {ex}")
ValueError: failed to update labels in query 'interface_rx_stats' with 'topology/pod-1/node-104/sys/ch/supslot-1/sup/cpuport-1/CDeqptIngrBytes5min': dictionary update sequence element #0 has length 1; 2 is required

[RavuAlHemio]

Ah yes, that's an issue with the configuration -- it expects a DN following the format

topology/pod-{POD}/node-{NODE}/sys/{PORTTYPE}-[{PORT}]/CDeqptIngrBytes5min

e.g.

topology/pod-1/node-101/sys/mgmt-[mgmt0]/CDeqptIngrBytes5min

or

topology/pod-1/node-101/sys/phys-[eth1/35]/CDeqptIngrBytes5min

but has instead obtained the DN

topology/pod-1/node-104/sys/ch/supslot-1/sup/cpuport-1/CDeqptIngrBytes5min

which doesn't follow this format, so it failed to extract the relevant values.

I'll look into integrating supslot/cpuport statistics into the sample configuration as well. Out of curiosity, what model is node 104 in your fabric? Is it a modular Nexus switch?

[insobi]

Node 104 is N9K-C9372PX. :) Anyway, like what you mentioned, retrieved dn format and expected dn format were not same. In this case, however, I supposed that it should be skipped without error if those are not same between retrieved and expected. Am I misunderstanding about the logic?