Add fields to info metricset of Redis module #7695
Conversation
|
Since this is a community submitted pull request, a Jenkins build has not been kicked off automatically. Can an Elastic organization member please verify the contents of this patch and then kick off a build manually? |
info metricset of Redis module| - name: used.rss | ||
| type: long | ||
| format: bytes | ||
| description: > | ||
| Used memory rss. | ||
|
|
||
| Number of bytes that Redis allocated as seen by the operating system (a.k.a resident set size). |
There was a problem hiding this comment.
Thanks for these improvements in the descriptions! 🙂
| "priority": c.Int("slave_priority", s.Optional), | ||
| "is_readonly": c.Bool("slave_read_only", s.Optional), | ||
| }, | ||
| // ToDo find a way to add dynamic object of slaves: "slaves": s.Str("slaveXXX") |
There was a problem hiding this comment.
Maybe slaves data could be in a specific metricset.
But something interesting we could add here are some aggregated metrics obtained from this list of slaves, like number of slaves in each state, average/min/max lag, min/max offset...
This can be left for another PR in any case.
There was a problem hiding this comment.
Here and in some other scenarios, I need a dynamic array in the schema. Is there any way to do so? Something like this JSON:
{
"slaves": [
{
"host": "localhost:6380",
"lag": 2
},
{
"host": "localhost:6381",
"lag": 0
},
]
}There was a problem hiding this comment.
This kind of structure can be problematic metric-wise, to make it useful, slaves should be a nested object, if not the event would be indexed as something like:
{
"slaves": {
"host": ["localhost:6380", "localhost:6381"]
"lag": [2, 0],
}
}
With this it is not possible, or very difficult, to make proper queries with filters or aggregations.
The problem with nested objects is that we are not using them in beats and they would require some additional support, as well as some performance tests. Kibana would also need better support for queries and visualizations on these kind of fields.
This is why I proposed to have an additional metricset, that could have events with fields like:
"slave": {
"host": "localhost:6380",
"lag": 2
}
Or to have by now aggregations directly on the existing metricset.
|
jenkins, test this |
|
@a3dho3yn you will need to update the branch and resolve the conflicts. |
| }, | ||
| }, | ||
| "slowlog": s.Object{ | ||
| "count": c.Int("slowlog_len"), |
There was a problem hiding this comment.
Don't forget to add this to the fields.yml file 🙂
|
I don't know why test_template (from test_base.py) fails :( |
| description: > | ||
| Ratio between used_memory_rss and used_memory | ||
| - name: fragmentation.bytes | ||
| type: bytes |
There was a problem hiding this comment.
@a3dho3yn I have run the failing test locally, and it seems that the failure is caused by this, bytes can be used as format, but not as type.
|
|
||
| - name: master_offset | ||
| type: long | ||
| description: | ||
|
|
||
| deprecated: true |
There was a problem hiding this comment.
After merging #7779 today, deprecated needs to be set to the version the field is going to be deprecated on, in this case set it to 6.5.
|
@a3dho3yn run |
|
I've run |
|
Oh yes, sorry, you also need to update your branch with master. |
|
jenkins, test this please |
|
@a3dho3yn merged, thanks a lot for this contribution and the others you are working on 😃 Just curious, are you using the provided dashboards for redis and rabbitmq? We'd be happy to accept contributions also there if you have any proposal 🙂 |
|
Thank you very much for all the help to complete these PRs. |
Add memory fragmentation ratio, active defragmentation stats, slave status and some other fields that were missing from info metricset. (cherry picked from commit ce9048c)
Memory fragmentation ratio, active defragmentation stats, slave status and some other fields was missing from
infometricset.