-
Notifications
You must be signed in to change notification settings - Fork 4.9k
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
introduce new indices
key with multi ilm and template handling
#10002
Conversation
02b3c3d
to
e4a76b1
Compare
os.Exit(1) | ||
} | ||
logp.Info("Loaded Elasticsearch templates.") | ||
}, |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
We will have to clean this up as well, so to not print all templates to stdout, right?
We print the templates, we do not load anything. Maybe we shouldn't print Loaded Elasticsearch templates.
once we are done.
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
We will have to clean this up as well, so to not print all templates to stdout, right?
We print exactly the templates that would be loaded - I think this behaves as expected.
We print the templates, we do not load anything. Maybe we shouldn't print Loaded Elasticsearch templates. once we are done.
agreed
return false | ||
} | ||
|
||
code, body, err := client.Request("GET", "/_xpack", "", nil, nil) |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
hm... how do we get the info once _xpack
is removed from the API?
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Good question, but this check is not introduced newly here. Can you link to the according issue where _xpack
should be removed, so I can follow up.
@urso I am closing this for now to avoid unnecessary noise, will reopen when it is final. |
libbeat/ilm/config.go
Outdated
*m = ModeAuto | ||
case "true", "True", "yes", "y", "Yes": | ||
*m = ModeEnabled | ||
case "false", "False", "no", "n", "No": |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
This is inconstant with the rest of beats we do not support all the boolean possibilities.
lets use an explicit "auto", "false", "true"
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
only lowercase or also uppercase?
@simitt I have still added a few notes to this PR, I will do a final review when its reopen. |
great, thanks for the feedback @ph ! |
Further changes and adaptions. Add metricbeat.yml as reference example.
662d763
to
f157a98
Compare
indices
key with multi ilm and template handlingindices
key with multi ilm and template handling
indices
key with multi ilm and template handlingindices
key with multi ilm and template handling
f157a98
to
81ebf66
Compare
Closing this PR as discussions around how to organise configuration settings properly in the future are still going on. |
Hi @simitt , is this work likely to be merged at any point? Not having this prevents us from being able to use ILM altogether as sending all of our cluster logs to the same index is introducing too many differing fields depending on the source of the log, leaving us with a Kibana that can't determine what fields are in our index patterns. We're disabling ILM for now in favor of using curator to do the same work for us but it looks like curator is being obsoleted over time in favor of ILM. |
Handing off this question to @urso, as the ILM related index handling for APM and beats developed differently. |
As discussed offline, this WIP PR is a start to address configuration changes related to
index, template, ilm
handling. Since this is a larger topic I created a feature branchfeature-multi-ilm
against which we can create PRs.related to ##9919
Points to cover:
indices
. Deprecate config optionssetup.template
,output.elasticsearch.index
,output.elasticsearch.indices
. Remove beta config optionilm
. See new config options example.indices
configuration if available. Currently anindex
is always set, even whenindices
are configured, as a fallback when no condition applies. In the future one index that doesn't have a condition must be set inindices
, which is used as the default index. If ILM is enabled, therollover_alias
must be used instead of thename
.template.name
andtemplate.pattern
be derived from therollover_alias
when ILM is enabled? Which fields are required to be set perindices
entry?@urso, @ph, @elastic/apm-server this PR is not ready for a detailed review, but I'd appreciate high-level feedback.
UPDATE:
indices
,output.elasticsearch.index
andsetup.template
sectionindex.DefaultConfig
setting so every beat can set its default index from outside. This is specifically helpful for apm-server (see Add indices defaults in code apm-server#1480). I am not exactly clear on what the default configurations for all the beats should be (especially with ECS changes). I'd appreciate if someone else could update the configs and overwrite theindex.DefaultConfig
if necessary. I tried to update the config formetricbeat
myself.