A performance degradation occurs on a gfs2 filesystem as the load increases on RHEL 6 or RHEL 7

Solution In Progress - Updated -

Issue

  • A write performance degradation occurs on a gfs2 filesystem as the load increases on RHEL 6 or RHEL 7.
  • When running the script called iotest.sh or rhel_iotest.sh (which are provided by SAS as benchmarking tool to test filesystem and storage) the write performance of a gfs2 filesystem decreased as the number of dd processes increased.
  • The same problem occurs when doing simultaneous dd commands or large simultaneous data writes using more CPU cores on the same node to the same filesystem.
  • This problem has nothing to do with the performance of the storage (though it still needs to be configured properly). It is about the rate a gfs2 filesystem can take write requests and place them into the page cache.
  • Performance seems to be capped at around 800-900 MB/sec per gfs2 filesystem.

Environment

  • Red Hat Enterprise Linux Server 6, 7 (with the High Availability and Resilient Storage Add Ons)
  • A Global Filesystem 2(GFS2)

Subscriber exclusive content

A Red Hat subscription provides unlimited access to our knowledgebase, tools, and much more.

Current Customers and Partners

Log in for full access

Log In

New to Red Hat?

Learn more about Red Hat subscriptions

Using a Red Hat product through a public cloud?

How to access this content