Skip to content
GitLab
Projects Groups Snippets
  • /
  • Help
    • Help
    • Support
    • Community forum
    • Submit feedback
    • Contribute to GitLab
  • Sign in / Register
  • BIND BIND
  • Project information
    • Project information
    • Activity
    • Labels
    • Members
  • Repository
    • Repository
    • Files
    • Commits
    • Branches
    • Tags
    • Contributors
    • Graph
    • Compare
  • Issues 570
    • Issues 570
    • List
    • Boards
    • Service Desk
    • Milestones
  • Merge requests 101
    • Merge requests 101
  • CI/CD
    • CI/CD
    • Pipelines
    • Jobs
    • Schedules
  • Deployments
    • Deployments
    • Environments
    • Releases
  • Packages and registries
    • Packages and registries
    • Container Registry
  • Monitor
    • Monitor
    • Incidents
  • Analytics
    • Analytics
    • Value stream
    • CI/CD
    • Repository
  • Wiki
    • Wiki
  • Snippets
    • Snippets
  • Activity
  • Graph
  • Create a new issue
  • Jobs
  • Commits
  • Issue Boards
Collapse sidebar
  • ISC Open Source ProjectsISC Open Source Projects
  • BINDBIND
  • Issues
  • #493
Closed
Open
Issue created Aug 23, 2018 by Cathy Almond@cathyaDeveloper

Measurable performance drop between BIND 9.10.6 and 9.10.7 [ISC-support #13354]

This was originally reported to ISC by a customer as a drop in performance between BIND 9.10 and 9.11, but later was confirmed to be seen also between 9.10.6-P1 and 9.10.7. BIND 9.11.3 and 9.11.4-P1 have similarly decreased performance. (This reporter has not tested BIND 9.11.2).

Test results can be found in https://support.isc.org/Ticket/Display.html?id=13354

All the most recent tests run with Debian 9 on the same platform: Dell PowerEdge R330 with 4 cores 3.0GHz, 16GB RAM. All 3 tests reported below are authoritative only, but the submitter says that they ran other scenarios with different HW and also recursive only and consistently observed a major drop in performance when running bind 9.11.3 versus BIND 9.10.6.

The test environment uses IXIA to simulate 100 clients generating DNS queries with a max QPS of 400K

The degradation in the most recent test set (9%) clearly starts in 9.10.7.

9.10.6-P1: 330 KQPS 9.10.7: 300 KQPS 9.11.4-P1: 300 KQPS

CPU is close to 100% in all tests. Available are top and ps outputs plus bind stats every 15 seconds during the tests.

The issue has also been seen in perflab: https://perflab.isc.org/#/config/run/56d4626b797da24b2741b1e9/

Opening this issue ticket at the request of engineering for them to home in on the underlying cause (or causes).

Assignee
Assign to
Time tracking