Trying to separate data from Line output in Grafana

Grafana Loki
, ,
1

I am creating a table panel in Grafana.

I am running a logql query in grafana:
{host=“hostname”}

The problem is the output is creating these separate fields: labels, Time, Line, tsNs, id. The “Line” column actually shows the output of syslog message. The output for “Line” field is as follows:
appname=xntp facility=ntp host=HOSTNAME message=blah…

I want to separate appname, facility, host, etc from the “Line” field, into their own separate fields. So, I can create a table dashboard with just the hostname and message for the columns.

How do I do this?

2

See pattern filter: LogQL: Log query language | Grafana Loki documentation

1 Like
3

I read through this documentation. I believe this is finding or not finding whatever pattern you query for. This is NOT what I’m trying to do. All the text in the output of Line, for example, is separated by a space. ie. appname=some_appname host=some_host etc. I want to take each block of text that is separated by a space and make a column for each block of text. So the table would look like this:

Column 1 Column 2 Column 3 Column 4 E F
appname host facility host message
app1 host_1 debug host_1 linkDown
app2 host_2 info host_2 linkUp
app3 host_3 info host_3 host_1_connected

Again, I only want everything extracted from the Line field. If I am incorrect in my thought process or this is included in the pattern documentation, please let me know. Thanks.

4

Actually, if your logs are consistently key=value separated by space, logfmt would be easier for you. See example: Simple LogQL simulator | Grafana Loki documentation

1 Like
5

Thank you for your help! The final fix is below for anyone reading this board. Goal: I wanted to only show certain fields in my output of log message for my dashboard and FILTER OUT ANYTHING ELSE:

logql query:
{host=“host1”} | logfmt | host != “” and message != “” | line_format “{{.host}} : {{.message}}”

6

Separating data from line output in Grafana can be done by using different visualization techniques, such as setting up different axes for distinct data sets or using transformations to manipulate the data before displaying it. If you’re looking for advanced tips, Koalageddon shares some great Grafana hacks. What kind of data are you trying to separate in your graph?

1 Like