[promtail helm chart] Enable support for syslog service #1617
Conversation
|
|
Codecov Report
@@ Coverage Diff @@
## master #1617 +/- ##
=======================================
Coverage 61.74% 61.74%
=======================================
Files 109 109
Lines 8308 8308
=======================================
Hits 5130 5130
Misses 2783 2783
Partials 395 395 |
billimek
changed the title
cyriltovena
requested review from
rfratto
and removed request for
sandeepsukhani
This should allow the chart to be configured for extra exposed ports like syslog (port 1514) for example Signed-off-by: Jeff Billimek <jeff@billimek.com>
Signed-off-by: Jeff Billimek <jeff@billimek.com>
Signed-off-by: Jeff Billimek <jeff@billimek.com>
Signed-off-by: Jeff Billimek <jeff@billimek.com>
|
Hi @rfratto, is there anything I can do with this PR to help it move along? |
There was a problem hiding this comment.
Hi @billimek, thank you for working on this! I'm not a Helm expert, so I'm afraid I can't give this the review it deserves. Just at a quick glance, my primary concern is that I don't see where it configures the syslog scrape config. Should that be included here?
@joe-elliott Could you take a look at this since you're more familiar with Helm?
|
Thanks @rfratto. As I understand it, the configuration of syslog would be handled in the extraScrapeConfigs section of the promtail helm chart configuration, e.g. extraScrapeConfigs:
- job_name: syslog
syslog:
listen_address: 0.0.0.0:1514
labels:
job: "syslog"
relabel_configs:
- source_labels: ['__syslog_message_hostname']
target_label: 'host'What's missing is having the promtail helm chart actually expose a syslog endpoint via a service - which is what this PR is attempting to solve. We could consider enhancing this pull request to encompass the entire syslog configuration. What do you think @joe-elliott? |
|
Looks good @billimek can you improve the documentation of the helm chart on how to use the syslog in promtail ? With the snippet above ? here https://github.com/grafana/loki/tree/master/production/helm Thanks ! |
Signed-off-by: Jeff Billimek <jeffrey_k_billimek@homedepot.com>
* 'extraPorts' of github.com:billimek/loki: (25 commits) Ensure status codes are set correctly in the frontend. (grafana#1684) --dry-run Promtail. (grafana#1652) Fix logcli --quiet parameter parsing issue (grafana#1682) This improves the log ouput for statistics in the logcli. (grafana#1644) Loki stack helm chart can deploy datasources without Grafana (grafana#1688) Automatically prune metrics from the /metrics output of the promtail metrics pipeline stage after an idle period. Allow a metric defined as a counter to match all lines, useful for creating line count metrics which include all your labels. Found a bug in the the test runner where it didn't fail if the actual error was nil but the test expected an error Added some tests to the counters to test valid configs maintainer links & usernames (grafana#1675) Binary operators in LogQL (grafana#1662) Pipe data to Promtail (grafana#1649) Add Owen to the maintainer team. (grafana#1673) Update tanka.md so that promtail.yml is in the correct format (grafana#1671) Correcte syntax of rate example (grafana#1641) Frontend & Querier query statistics instrumentation. (grafana#1661) loki-canary: fix indent of DaemonSet manifest written in .md file (grafana#1648) Query frontend service should be headless. (grafana#1665) Support custom prefix name in metrics stage (grafana#1664) pkg/promtail/positions: handle empty positions file (grafana#1660) Convert second(Integer class) to nanosecond precision (grafana#1656) Unite docs for fluentd plugin (grafana#1634) ...
Thanks @cyriltovena - this should be present now. Can you please take a look? |
What this PR does / why we need it:
This should allow the chart to be configured for an optional syslog service
Special notes for your reviewer:
Checklist