Billing module delta issue after device recovers from outage

  • Steps to reproduce an issue.
  • Billing monitor of a device that has semi-regular drop outs
  • Devices being monitored specifically are Mikrotik RouterBoard of various firmware

We have an issue where the usage delta is massive after a monitored device’s connection is restored after an outage. In our billing module we are using quota not CDR.

I have checked the SQL data and have found that when the device is unreachable it is recording a 0 for that check-in rather than what I would expect to be NaN to avoid the delta split between 0 and it’s current counter.

Is this behaviour expected, or is there a setting we may have misconfigured?
See below for SQL export and a custom SNMP walk for the OID

+---------+---------------------+--------+-----------+-----------+-----------+
| bill_id | timestamp | period | delta | in_delta | out_delta |
+---------+---------------------+--------+-----------+-----------+-----------+
| 5 | 2023-08-21 09:25:03 | 301 | 57151 | 24517 | 32634 |
| 5 | 2023-08-21 09:20:02 | 299 | 58237 | 24903 | 33334 |
| 5 | 2023-08-21 09:15:03 | 1 | 760 | 380 | 380 |
| 5 | 2023-08-21 09:15:02 | 300 | 56611 | 24247 | 32364 |
| 5 | 2023-08-21 09:10:03 | 300 | 583675333 | 222498727 | 361176606 |
| 5 | 2023-08-21 09:05:03 | 300 | 0 | 0 | 0 |
| 5 | 2023-08-21 09:00:03 | 300 | 0 | 0 | 0 |
| 5 | 2023-08-21 08:55:03 | 300 | 0 | 0 | 0 |
| 5 | 2023-08-21 08:50:03 | 301 | 0 | 0 | 0 |
| 5 | 2023-08-21 08:45:02 | 299 | 0 | 0 | 0 |
| 5 | 2023-08-21 08:40:03 | 325 | 115426 | 41722 | 73704 |
| 5 | 2023-08-21 08:35:02 | 299 | 115426 | 41722 | 73704 |
| 5 | 2023-08-21 08:30:03 | 300 | 57328 | 22030 | 35298 |
| 5 | 2023-08-21 08:25:03 | 300 | 57681 | 25067 | 32614 |
| 5 | 2023-08-21 08:20:03 | 300 | 57823 | 24853 | 32970 |
| 5 | 2023-08-21 08:15:03 | 300 | 57151 | 24517 | 32634 |
| 5 | 2023-08-21 08:10:03 | 301 | 59319 | 25471 | 33848 |
| 5 | 2023-08-21 08:05:02 | 299 | 56231 | 24057 | 32174 |
| 5 | 2023-08-21 08:00:03 | 300 | 57703 | 24793 | 32910 |

And the SNMP walk with the OID counter current value (no delta calculation done).

20230821-08:54   000.207179 GB           000.336307 GB
20230821-08:55
20230821-08:56
20230821-08:57
20230821-08:58
20230821-08:59
20230821-09:00
20230821-09:01
20230821-09:02   000.207187 GB           000.336332 GB
20230821-09:03   000.207190 GB           000.336335 GB
20230821-09:04   000.207192 GB           000.336338 GB
20230821-09:05   000.207194 GB           000.336341 GB
20230821-09:06   000.207208 GB           000.336360 GB
20230821-09:07   000.207210 GB           000.336362 GB
20230821-09:08   000.207212 GB           000.336365 GB
20230821-09:09   000.207215 GB           000.336368 GB
20230821-09:10   000.207217 GB           000.336371 GB
20230821-09:11   000.207230 GB           000.336389 GB

Bump. Has anyone run into this issue or have and insights that may be helpful?

This topic was automatically closed 90 days after the last reply. New replies are no longer allowed.