[Bug]: Chart filtering to prometheus appears to be broken about netdata

[ilyam8]

Hi, @johnalotoski. I can't reproduce the problem in v1.45.4.

$ cat exporting.conf | grep -v "#"
[prometheus:exporter]
    send charts matching = !*

$ curl -s 'localhost:19999/api/v1/allmetrics?format=prometheus' | wc -l
1

[ilyam8]

Have you restarted Netdata after updating exporting.conf?

Also, you can filter using the filter URL parameter:

$ curl 'localhost:19999/api/v1/allmetrics?format=prometheus&filter=*system.softirq*'
netdata_info{instance="pve-deb-work",application="netdata",version="v1.45.3-10-gb589731c8"} 1 1715237847219
netdata_system_softirq_latency_milliseconds_average{chart="system.softirq_latency",dimension="HI",family="softirqs"} 0.0000000 1715237830000
netdata_system_softirq_latency_milliseconds_average{chart="system.softirq_latency",dimension="TIMER",family="softirqs"} 0.2000000 1715237830000
netdata_system_softirq_latency_milliseconds_average{chart="system.softirq_latency",dimension="NET_TX",family="softirqs"} 0.0000000 1715237830000
netdata_system_softirq_latency_milliseconds_average{chart="system.softirq_latency",dimension="NET_RX",family="softirqs"} 0.0000000 1715237830000
netdata_system_softirq_latency_milliseconds_average{chart="system.softirq_latency",dimension="BLOCK",family="softirqs"} 0.0000000 1715237830000
netdata_system_softirq_latency_milliseconds_average{chart="system.softirq_latency",dimension="IRQ_POLL",family="softirqs"} 0.0000000 1715237830000
netdata_system_softirq_latency_milliseconds_average{chart="system.softirq_latency",dimension="TASKLET",family="softirqs"} 0.0000000 1715237830000
netdata_system_softirq_latency_milliseconds_average{chart="system.softirq_latency",dimension="SCHED",family="softirqs"} 0.4000001 1715237830000
netdata_system_softirq_latency_milliseconds_average{chart="system.softirq_latency",dimension="HRTIMER",family="softirqs"} 0.0000000 1715237830000
netdata_system_softirq_latency_milliseconds_average{chart="system.softirq_latency",dimension="RCU",family="softirqs"} 0.4000001 1715237830000
netdata_system_softirqs_softirqs_persec_average{chart="system.softirqs",dimension="HI",family="softirqs"} 0.0000000 1715237844000
netdata_system_softirqs_softirqs_persec_average{chart="system.softirqs",dimension="TIMER",family="softirqs"} 53.1620671 1715237844000
netdata_system_softirqs_softirqs_persec_average{chart="system.softirqs",dimension="NET_TX",family="softirqs"} 0.0000000 1715237844000
netdata_system_softirqs_softirqs_persec_average{chart="system.softirqs",dimension="NET_RX",family="softirqs"} 12.3700883 1715237844000
netdata_system_softirqs_softirqs_persec_average{chart="system.softirqs",dimension="BLOCK",family="softirqs"} 0.5258970 1715237844000
netdata_system_softirqs_softirqs_persec_average{chart="system.softirqs",dimension="TASKLET",family="softirqs"} 0.1428571 1715237844000
netdata_system_softirqs_softirqs_persec_average{chart="system.softirqs",dimension="SCHED",family="softirqs"} 69.8122657 1715237844000
netdata_system_softirqs_softirqs_persec_average{chart="system.softirqs",dimension="HRTIMER",family="softirqs"} 0.0000000 1715237844000
netdata_system_softirqs_softirqs_persec_average{chart="system.softirqs",dimension="RCU",family="softirqs"} 137.9602871 1715237844000
$

[johnalotoski]

Hi @ilyam8, thanks for trying to reproduce; I do have this config issue resolved on my side now.

For filtering by URL parameter, with the example you provided, I did address that above in the Additional Info section -- see the log example there. Some common scrape clients, such as grafana-agent, escape the pattern matchers, such as * in the URL, so setting params for filtering through these clients doesn't work because of the automatic URL escaping. I wish it did work there too.