https://signoz.io logo
Join SlackCommunities
Powered by
This message was deleted.
# support
s
This message was deleted.
s
No, the otel-collector acts like a prometheus server
k
does it support all stuff like consul sd_config which are in promethues
can we replace complete prometheus with otel collector and signoz query service
just correct me if i am wrong i have prometheus node exporter and apache exporter is that gonna directly support in otel collector
s
It's not a drop-in replacement but it will be possible to set up the otel collector to collect all your Prometheus metrics.
What other components do you have that emit the Prometheus metrics other than node exporter and apache.
k
netapp exporter
cadvisor exporter
consul exporter
there are lot
so how to handle this situtation
like promethues is running and can we change the datasource to write to clikchouse
or it need to be go via otel collector only into clickhouse
s
Like I said there is no drop in replacement. There needs to be some component that writes to ClickHouse which is the otel collector. Roughly how many do you have? Are these components running inside the k8s cluster and annotated 
<http://promethues.io/scarpe|promethues.io/scarpe>: true
etc..? If so they can be automatically collected with service discovery.
k
no its monollithic with non k8s but docker based and non docker based
s
What does your infrastructure look like and where do these components run?
k
its vm based application
in-premise
i am reading your documentation
in that enable prometheus recevier
will fetch all data from prometheus and write to clickhouse
its like scraping from prometheus and ingesting into clickhouse
does take it take all scrape config and metrics from prometheus and store it
?
rather than configuring receiver can't we add remote write into clickhouse itself ?
from prometheus instead of otel-collector
s
Repeating myself if you were not able to follow earlier comments. No, it's not possible. There must be some component that translates the Prometheus protocol/format data and write to ClickHouse, which is the otel collector.
its like scraping from prometheus and ingesting into clickhouse
It's not scraping from prometheus. It is scraping from components that expose prometheus metrics.
does take it take all scrape config and metrics from prometheus and store it
To clarify again, it's not going to interact with Prometheus server. The collector mimics the Promethus server behaviour to scrape data.
k
k thanks
whats your thoughts on this
s
This is what we have been discussing all along and the component highlighted in our docs (Enable a Prometheus Receiver).
k
i am able to get the metrics from node_exporter by adding into receiver
just curious so when prometheus receiver receving data how much of data it can receive or where it is storing this metrics ?
its reading from metrics from node exporter which is installed on machine and otel collector is scraping the data with the help of defination pointing on the node itself which is deployed on node itself
and now all node which has defination of prometheus receiver in all collector will send data to signoz collector which is centrally located
correct me if my workflow is wrong
s
The receiver is only responsible for reading the metrics and sending it to the next pipeline. The data ultimately gets stored in ClickHouse. You don't have to deploy a collector in all nodes. You can create a node which has access to all other nodes and just deploy collector on this node which scrapes metrics from all configured targets.
k
perfect thanks for clarifying
i wish if we can import grafana dashboard
so it should be all in one tool
not to use all tools and use signoz
only
3 Views
Open in Slack
PreviousNext
Powered by