Uploaded image for project: 'Couchbase Server'
  1. Couchbase Server
  2. MB-45843

Investigate space cost of KV stats

    XMLWordPrintable

Details

    • Task
    • Resolution: Done
    • Major
    • None
    • Cheshire-Cat
    • couchbase-bucket
    • None
    • 1

    Description

      KV maintains a total of 388 stats.
      Some of these stats have high cardinality, such as those that produce histograms.

      We want to establish..

      • what is the space usage is in terms of MB / hour?
      • which stats are responsible the majority of consuming the large space?
      • How many high-cardinality stats do we maintain?
      • Do we really need 388 stats? Are some redundant / of very limited value?

      Approach

      • Set-up a single KV node cluster and monitor disk usage over a period of 24 hours
      • Repeat with a constant KV low-load

      Attachments

        1. 30-bucket-CPU.png
          30-bucket-CPU.png
          14 kB
        2. 30-bucket-RAM.png
          30-bucket-RAM.png
          32 kB
        3. counts.xlsx
          1.14 MB
        4. Dashboard-stats.png
          Dashboard-stats.png
          38 kB
        5. diskspace.png
          diskspace.png
          16 kB
        6. growth.csv
          35 kB
        7. image.png
          image.png
          14 kB
        8. output.txt
          166 kB
        9. promtimer-limited-data.png
          promtimer-limited-data.png
          69 kB
        10. Screenshot 2021-05-21 at 7.48.01 PM.png
          Screenshot 2021-05-21 at 7.48.01 PM.png
          257 kB
        11. tsdb.txt
          3 kB

        Issue Links

          relates to

          Bug - A problem which impairs or prevents the functions of the product. MB-46046 All histogram buckets should be reported in every collection

          • Major - Major loss of function.
          • Closed

          Bug - A problem which impairs or prevents the functions of the product. MB-47153 Use /_prometheusMetrics*NoTS endpoints for external prometheus use case (for kv only)

          • Major - Major loss of function.
          • Closed

          Epic - Created by Jira Software - do not edit or delete. Issue type for a big user story that needs to be broken down. MB-47091 Extend Stats Coverage Period

          • Critical - Crashes, loss of data, severe memory leak.
          • Open

          Task - A task that needs to be done. MB-46422 Remove kv_timings histograms from being stored in Prometheus

          • Major - Major loss of function.
          • Resolved
          No reviews matched the request. Check your Options in the drop-down menu of this sections header.

          Activity

            People

              owend Daniel Owen
              owend Daniel Owen
              Votes:
              0 Vote for this issue
              Watchers:
              10 Start watching this issue

              Dates

                Created:
                Updated:
                Resolved:

                Gerrit Reviews

                  There are no open Gerrit changes

                  PagerDuty