[[inputs.win_perf_counters.object]]
Using the above configuration, the telegraf sends two message with one counter each. I expected one message with two counter values.
win_disk,host=WIN-GVVTNU0V3TV,item_identifier=2709:2569,category=agility-vm,interval=30s,instance=C:,objectname=LogicalDisk Free_Megabytes=45317 1494533452000000000
win_disk,interval=30s,host=WIN-GVVTNU0V3TV,item_identifier=2709:2569,category=agility-vm,instance=C:,objectname=LogicalDisk Percent_Free_Space=74.18435668945312 1494533485000000000
Anyone using this and seeing this?
YES!
I thought I was losing my mind with this. For every collection interval I see only one counter collected within each inputs.win_perf_counters.object. This behavior is really bad with the Physical Disk Object with its 8 default and useful counters. I cant correlate spikes in queue length with read/write IO or rates.
Should telegraf be collecting all of them with each interval? Or is there a setting I am missing. I am running Influx and Grafana on CentOS 7, telegraf on Windows Server 2012. I have tested and confirmed this behavior on v1.3.1 along with 1.0.0 beta3.
@bj2001holt This sounds like an issue. Can you open an issue on telegraf? Sorry for not getting back sooner, I’m not a Windows user so I can’t really answer!
bj2001holt:
1.3.1
Opened the issue, thanks!
opened 06:33PM - 07 Jun 17 UTC
closed 12:48AM - 01 May 18 UTC
bug
platform/windows
area/windows
## Bug report
When using an "inputs.win_perf_counters.object" where you are col… lecting multiple counters you will only see a single counter collected at each interval. For example, the PhysicalDisk Object collects 8 counters, however it only collects one of them per interval. This removes the ability to correlate performance data.
### Relevant telegraf.conf:
[agent]
interval = "5s"
round_interval = true
metric_buffer_limit = 1000
flush_buffer_when_full = true
flush_interval = "5s"
flush_jitter = "0s"
[[outputs.influxdb]]
precision = "s"
[[inputs.win_perf_counters.object]]
ObjectName = "PhysicalDisk"
Instances = ["*"]
Counters = [
"Disk Read Bytes/sec",
"Disk Write Bytes/sec",
"Current Disk Queue Length",
"Disk Reads/sec",
"Disk Writes/sec",
"% Disk Time",
"% Disk Read Time",
"% Disk Write Time",
]
Measurement = "win_diskio"
### System info:
Influx and Grafana on CentOS 7, telegraf on Windows Server 2012. Tested with telegraf version 1.3.1 and 1.0.0 beta 3.
### Steps to reproduce:
1. Install telegraf on windows system using default win_perf_counters
2. From Influx query
SELECT Current_Disk_Queue_Length as "C: Current Queue Length", Disk_Reads_persec as "C: Read/Sec", Disk_Writes_persec as "C: Writes/Sec" FROM "win_diskio" WHERE "host" = '<removed>' AND "instance" = '0 C:'
### Expected behavior:
All counters should be collected with every interval
### Actual behavior:
One counter is collected in each interval
1 Like
