Updated ScyllaDB Prometheus dashboard and documentation #588
Conversation
There was a problem hiding this comment.
Flagging to @EvanSimpson as a second breaking change case, even though the first such case (#581 for postgres) may be a false alarm
| @@ -4,7 +4,7 @@ platforms: | |||
| exporter_metadata: | |||
| name: ScyllaDB Prometheus Exporter | |||
| doc_url: https://monitoring.docs.scylladb.com/stable/reference/monitoring_apis.html | |||
| minimum_supported_version: "5.0" | |||
| minimum_supported_version: "5.2" | |||
| detections: | |||
| - characteristic_metric: | |||
There was a problem hiding this comment.
We'll need to update the list of default metrics below too
| "scale": "LINEAR" | ||
| } | ||
| } | ||
| }, | ||
| "width": 12, | ||
| "xPos": 0, | ||
| "yPos": 28 | ||
| "yPos": 7 |
There was a problem hiding this comment.
Does this mean we have to retake the screenshot preview of the dashboard in the same folder?
Does the dashboard currently query the metrics which are going away? If the dashboard is fully backwards-compatible with both 5.0 and 5.2 metrics, I think we are good to bump the version and update the list of default metrics to match what 5.2 collects. If not, we should consider a new integration version with a forked dashboard. |
|
@EvanSimpson Apologies for not answering sooner, the dashboard does currently query those metrics that are going away. I believe that means it won't be backwards-compatible with both versions. I think what you've suggested will be the best path forward. |
* updated min exporter version to be the latest version of Velero supported by the GCP plugin for Velero * the plugin support matrix doesn't match what is deployed, changed the minimum supported version * updated the minimum exporter version to be 1.7.1
* removed args for the exporter due to them all being default values, updated min version * updated the args to include web.listen-address which replaced telemetry.address * updated launch_stage to GA for apache http metadata
…tional install info
* removed deprecated metric and updated detection metric * removed deprecated metric from chart, updated screenshot
* Updated haproxy config, port name, and image version. Updated minimum version to match haproxy version * Changed haproxy exporter type to be 'included' and updated the documentation to fit that template * added back deployment yaml, named port, and metric verification instructions * added class to links, added new lines * small wording changes
…es based on the content in ops-agent repo.
Modify project to project_id
* updated the minimum exporter version, example configuration, and additional install info * updated the minimum supported version in metadata * update config example to pass a config file as an argument * reverted min versions, added volume to the config example that mounts the mysql exporter config * added + symbols for configmap in example yaml * removed a few dashes, updated the volume section, added mysql image to container section
|
@EvanSimpson I've pushed updates to version the dashboards, please let me know if there are any changes necessary. Hope you're having a great day! |
|
@algchoo thanks! It looks like there are a lot of unrelated changes in this PR, does this branch need to be rebased? |
This comment was marked as spam.
This comment was marked as spam.
…es based on the content in ops-agent repo.
…es based on the content in ops-agent repo.
…es based on the content in ops-agent repo.
I've gone ahead and rebased after making sure our fork was up to date, it should be good to go now. Hope you're having a good day! |
|
@EvanSimpson I still see a handful of changes that weren't part of what I was doing here, maybe I rebased incorrectly? On the plus side, there are no longer conflicts. |
I tried the quick fix described here to change the target to something else and then back but it doesn't seem to have worked https://stackoverflow.com/questions/16306012/github-pull-request-showing-commits-that-are-already-in-target-branch Can you try one of the options outlined in this answer? https://stackoverflow.com/a/45286781 |
|
@EvanSimpson I've opened a new PR to see if that helps solve this problem, here's a link and I'll close this PR. |
|
Wtf is this and why would I get an invoice for 596+
Most Likely Sent from Someone, Somewhere on a device or via smoke signal’s
from somewhere over the rainbow
…On Thu, Aug 24, 2023 at 2:08 PM Austin Georgiades ***@***.***> wrote:
Closed #588
<#588>
.
—
Reply to this email directly, view it on GitHub
<#588 (comment)>,
or unsubscribe
<https://github.com/notifications/unsubscribe-auth/AYHSMFCZXKMJBBWWD3GER6TXW6RELANCNFSM6AAAAAA2KWTGHQ>
.
You are receiving this because you commented.Message ID:
<GoogleCloudPlatform/monitoring-dashboard-samples/pull/588/issue_event/10185082066
@github.com>
|
|
I have emailed you in the past telling you that I had been hacked and that
I have no idea why I am getting your emails. I will be going to the
authorities
Most Likely Sent from Someone, Somewhere on a device or via smoke signal’s
from somewhere over the rainbow
On Fri, Sep 15, 2023 at 1:59 PM Robin Turner ***@***.***>
wrote:
… Wtf is this and why would I get an invoice for 596+
Most Likely Sent from Someone, Somewhere on a device or via smoke signal’s
from somewhere over the rainbow
On Thu, Aug 24, 2023 at 2:08 PM Austin Georgiades <
***@***.***> wrote:
> Closed #588
> <#588>
> .
>
> —
> Reply to this email directly, view it on GitHub
> <#588 (comment)>,
> or unsubscribe
> <https://github.com/notifications/unsubscribe-auth/AYHSMFCZXKMJBBWWD3GER6TXW6RELANCNFSM6AAAAAA2KWTGHQ>
> .
> You are receiving this because you commented.Message ID:
> <GoogleCloudPlatform/monitoring-dashboard-samples/pull/588/issue_event/10185082066
> @github.com>
>
|
Changes
Details
In version 5.2 of ScyllaDB, the following metrics are no longer present
scylla_storage_proxy_coordinator_writes_coordinator_outside_replica_set(Write Requests)scylla_storage_proxy_coordinator_write_timeouts(Write Timeouts)scylla_storage_proxy_coordinator_read_timeouts(Read Timeouts)The Read Performance and Write Performance both have the timeout metrics removed, while an equivalent has replaced the
scylla_storage_proxy_coordinator_writes_coordinator_outside_replica_set(Write Requests) metric:The minimum version has been updated to
5.2in the documentation and the metadata.