Grafana query drop downs not having all metrics

I used to be seeing in this drop down all possible options, but I am not seeing these any more. Is this even an issue of grafana or could this be because of an update in influxdb?

using grafana-4.6.3-1.x86_64 on CentOS6 (default repo baseurl=https://packagecloud.io/grafana/stable/el/6/$basearch)

Normally this list would contain lots of entries, maybe 200+ like these
type_instance=ThrottleMsgrDispatchThrottlerCluster.val
type_instance=ThrottleMsgrDispatchThrottlerHbBackClient.max
type_instance=ThrottleMsgrDispatchThrottlerHbBackClient.val
type_instance=ThrottleMsgrDispatchThrottlerHbBackServer.max
type_instance=ThrottleMsgrDispatchThrottlerHbBackServer.val
type_instance=ThrottleMsgrDispatchThrottlerHbclient.max
type_instance=ThrottleMsgrDispatchThrottlerHbclient.val
type_instance=ThrottleMsgrDispatchThrottlerHbFrontClient.max
type_instance=ThrottleMsgrDispatchThrottlerHbFrontClient.val
type_instance=ThrottleMsgrDispatchThrottlerHbFrontServer.max
type_instance=ThrottleMsgrDispatchThrottlerHbFrontServer.val
type_instance=ThrottleMsgrDispatchThrottlerMds.max
type_instance=ThrottleMsgrDispatchThrottlerMds.val
type_instance=ThrottleMsgrDispatchThrottlerMgr0x5603eea71ae0.max
type_instance=ThrottleMsgrDispatchThrottlerMgr0x5603eea71ae0.val
type_instance=ThrottleMsgrDispatchThrottlerMgr.max
type_instance=ThrottleMsgrDispatchThrottlerMgr.val
type_instance=ThrottleMsgrDispatchThrottlerMon.max
type_instance=ThrottleMsgrDispatchThrottlerMonMgrc.max
type_instance=ThrottleMsgrDispatchThrottlerMonMgrc.val
type_instance=ThrottleMsgrDispatchThrottlerMon.val
type_instance=ThrottleMsgrDispatchThrottlerMsObjecter.max
type_instance=ThrottleMsgrDispatchThrottlerMsObjecter.val
type_instance=ThrottleObjecterBytes.max
type_instance=ThrottleObjecterBytes.val
type_instance=ThrottleObjecterOps.max
type_instance=ThrottleObjecterOps.val
type_instance=ThrottleOsdClientBytes.max
type_instance=ThrottleOsdClientBytes.val
type_instance=ThrottleOsdClientMessages.max
type_instance=ThrottleOsdClientMessages.val
type_instance=ThrottleWriteBufThrottle0x561374596120.max
type_instance=ThrottleWriteBufThrottle0x561374596120.val
type_instance=ThrottleWriteBufThrottle.max
type_instance=ThrottleWriteBufThrottle.val
type_instance=WBThrottle.bytesDirtied
type_instance=WBThrottle.bytesWb
type_instance=WBThrottle.inodesDirtied
type_instance=WBThrottle.inodesWb
type_instance=WBThrottle.iosDirtied
type_instance=WBThrottle.iosWb

Hi,

What version of Grafana do you use? I can confirm that it works in Grafana v5.0 beta 2, see example panel.

Marcus

I am using the CentOS6 grafana repo currently at
grafana-4.6.3-1.x86_64

Strange thing is that I used to have this, and all of sudden it is gone. Maybe it is in some cache that needs to be rebuild or maybe this is because of influxdb-1.4.2-1.x86_64 (CentOS7)

Hi,

Thanks. Please check chrome dev tools -> network tab when you expand the drop down and inspect the request/response of the http request. Maybe it’s taking some time for that query to complete and/or it timing out.

Marcus

Yes good point, looks like I have all 200 and 304 status, total load 872ms, console also doesn’t list anything. All connections to the grafana server (that is acting as proxy to the influxdb)

Think I maybe know why you not receiving what you expect. All earlier filters are used when fetching the list for the drop down so I can see that you have a type and hostname filter applied. So if you just try and remove all WHERE filters and add a new one I’m guessing you should see more things.

Marcus

I wish this were the case, but I have already been modifying existing queries, creating new queries, even new dashboards, different datasource etc
If I remove everything after where and click the + no drop down, I think I used to be able to choose there type / type_instance or so.

Too bad. Let’s see what request/response I get when I select a tag and then a value. And then maybe you can provide the same?

  1. Open tag drop down

request:

response:

{
  "results": [
    {
      "statement_id": 0,
      "series": [
        { "name": "logins.count", "columns": ["tagKey"], "values": [["datacenter"], ["hostname"], ["source"]] }
      ]
    }
  ]
}
  1. Select tag hostname

request:
Query String Parameters:

db:site
q:SELECT mean("value") FROM "logins.count" WHERE ("hostname" = 'select tag value') AND time >= now() - 30m GROUP BY time(10m), "hostname";SELECT mean("value") FROM "logins.count" WHERE time >= now() - 30m GROUP BY time(10m)
epoch:ms

response:

{
  "results": [
    { "statement_id": 0 },
    {
      "statement_id": 1,
      "series": [
        {
          "name": "logins.count",
          "columns": ["time", "mean"],
          "values": [
            [1519718400000, null],
            [1519719000000, null],
            [1519719600000, 150.08823271144462],
            [1519720200000, 153.42556716683072]
          ]
        }
      ]
    }
  ]
}
  1. Open tag value drop down

request:
Query String Parameters:

db:site
q:SHOW TAG VALUES FROM "logins.count" WITH KEY = "hostname"
epoch:ms

response:

{
  "results": [
    {
      "statement_id": 0,
      "series": [
        {
          "name": "logins.count",
          "columns": ["key", "value"],
          "values": [
            ["hostname", "10.1.100.1"],
            ["hostname", "10.1.100.10"],
            ["hostname", "asd/ space/ metric"],
            ["hostname", "asd2\\\\ asd\\\\1"],
            ["hostname", "server/7"],
            ["hostname", "server1"],
            ["hostname", "server2"],
            ["hostname", "server3"],
            ["hostname", "server4"],
            ["hostname", "server\\\\5"]
          ]
        }
      ]
    }
  ]
}
  1. Select tag value server/7

request:
Query String Parameters:

db:site
q:SELECT mean("value") FROM "logins.count" WHERE ("hostname" = 'server/7') AND time >= now() - 30m GROUP BY time(10m), "hostname";SELECT mean("value") FROM "logins.count" WHERE time >= now() - 30m GROUP BY time(10m)
epoch:ms

response:

{
  "results": [
    {
      "statement_id": 0,
      "series": [
        {
          "name": "logins.count",
          "tags": { "hostname": "server/7" },
          "columns": ["time", "mean"],
          "values": [
            [1519718400000, null],
            [1519719000000, null],
            [1519719600000, 149.98229276817585],
            [1519720200000, 152.9506067740172]
          ]
        }
      ]
    },
    {
      "statement_id": 1,
      "series": [
        {
          "name": "logins.count",
          "columns": ["time", "mean"],
          "values": [
            [1519718400000, null],
            [1519719000000, null],
            [1519719600000, 150.08823271144462],
            [1519720200000, 153.07606812517076]
          ]
        }
      ]
    }
  ]
}

Marcus

Nice to learn this, this url is getting a correct overview of measurements
http://192.168.11.5:3000/api/datasources/proxy/1/query?db=collections&q=SHOW%20MEASUREMENTS%20LIMIT%20100&epoch=ms

This is getting
http://192.168.11.5:3000/api/datasources/proxy/1/query?db=collections&q=SHOW%20TAG%20KEYS%20FROM%20"ceph_value"&epoch=ms

{“results”:[{“statement_id”:0}]}

I tried the query “show tag keys from ceph_value;” on the influxdb console where it also does not return any values. So I think I better post this then on the influxdb community not? Or have you seen this happen before?

show tag keys from disk_value; is also not returning anything

1 Like

Looks like influxdb issue, looks like I am not the only one having this,

1 Like