paregupt/ucs_traffic_monitor

Missing measurements tables

Closed this issue · 4 comments

Hi,
I installed the OVA yesterday and tried the scripte with two Domains. It was working fine.
Then I added all my 7 domains in 7 locations. Also I changed from IP Address to Domainname.

To have only hostnames in the selection dropdown, I followed the Instructions and droped the telegraf databese.
The DB is there but there are measurements tables midding.

`> show measurements
name: measurements
name

cpu
disk
diskio
kernel
mem
net
processes
swap
system`

So I restarted the whole setup with the OVA and added all my domains from the Beginning. But still the tabeles are misssing.
Is there a Limit for the number of Domains?
Is it a must to user IPs for the Domain?

The Log as on error, see below.

/usr/local/telegraf/ucs_domains_group_1.txt:

[LAN]
10.97.100.30,ucs-local\user,pwd

LOG:

`....

2021-09-22 10:16:28,362 - ERROR - Exception with print_output:
'transport'
Traceback (most recent call last):
File "/usr/local/telegraf/ucs_traffic_monitor.py", line 2676, in main
print_output()
File "/usr/local/telegraf/ucs_traffic_monitor.py", line 2631, in print_output
print_output_in_influxdb_lp()
File "/usr/local/telegraf/ucs_traffic_monitor.py", line 2349, in print_output_in_influxdb_lp
',transport=' + per_fi_port_dict['transport']
KeyError: 'transport'
2021-09-22 10:16:28,363 - INFO - Pickle sessions for next time in /usr/local/telegraf/ucs_traffic_monitor_ucs_domains_group_1.pickle : {'10.97.100.30': {'sdk': <ucsmsdk.ucshandle.UcsHandle object at 0x7efd251a1198>, 'sdk_time': 1632294844, 'cli': None}, '10.97.100.40': {'sdk': <ucsmsdk.ucshandle.UcsHandle object at 0x7efd2515e6d8>, 'sdk_time': 1632294859, 'cli': None}, '10.97.100.10': {'sdk': <ucsmsdk.ucshandle.UcsHandle object at 0x7efd251665f8>, 'sdk_time': 1632294846, 'cli': None}, '10.97.100.20': {'sdk': <ucsmsdk.ucshandle.UcsHandle object at 0x7efd251660b8>, 'sdk_time': 1632294856, 'cli': None}, '10.97.100.60': {'sdk': <ucsmsdk.ucshandle.UcsHandle object at 0x7efd2516ccc0>, 'sdk_time': 1632294845, 'cli': None}, '10.97.100.70': {'sdk': <ucsmsdk.ucshandle.UcsHandle object at 0x7efd2516cac8>, 'sdk_time': 1632294844, 'cli': None}, '10.97.100.80': {'sdk': <ucsmsdk.ucshandle.UcsHandle object at 0x7efd2516ca58>, 'sdk_time': 1632294843, 'cli': None}}
2021-09-22 10:16:28,364 - INFO -
...
|--------------------------------------------|
| Response time - 10.97.100.30 |
|--------------------------------------------|
| CLI: Login: 5.6 s | Query: 10.09 s |
| SDK: Login: 0.62 s | Query: 9.79 s |
| Total: 15.69 s |
|--------------------------------------------|`

It looks like, the script has problems with ucs-mini.
I removed it from the ucs_domains_group_1, now the tables exist and I can see them in the graph.

Make sure you upgrade to the latest. UCS mini support is unavailable in the OVA version.
Split the domains in multiple files. 7 domains in the same file may not scale.
Check #63 and #64 for help on cleaning data.
Many others have run into the similar issues and found the solutions. The existing issues (open and closed) can give you detailed insights.

@ZenHub01 did it work?

See this to scale up UTM: https://www.since2k7.com/blog/2020/02/29/cisco-ucs-monitoring-using-grafana-influxdb-telegraf-utm-installation/#comment-55
Closing this issue. Feel free to open this issue if the problem persists. For a newer problem, feel free to open a new issue.