Bug 1387592

With this release, metadata cache on the client side is integrated with the cache-invalidation feature so that the clients can cache longer without compromising on consistency.
    By enabling, the metadata cache and cache invalidation feature and extending the cache timeout to 600s, we have seen performance improvements in metadata operation like creates, ls/stat, chmod, rename, delete.
    The perf improvements is significant in SMB access of gluster volume, but as a cascading effect the improvements is also seen on FUSE/Native access and NFS access.

Use the below options in the order mentioned, to enable the features:
  # gluster volume set <volname> features.cache-invalidation on
  # gluster volume set <volname> features.cache-invalidation-timeout 600
  # gluster volume set <volname> performance.stat-prefetch on
  # gluster volume set <volname> performance.cache-invalidation on
  # gluster volume set <volname> performance.cache-samba-metadata on     #Need to be set only for SMB access
  # gluster volume set <volname> performance.md-cache-timeout 600

Hi Poornima,

As discussed, I have created a new section named "Tuning Performance":

http://ccs-jenkins.gsslab.brq.redhat.com:8080/job/doc-Red_Hat_Gluster_Storage-3.2-Administration_Guide-branch-bz-1374150-gluster_nfs_off_by_default/lastSuccessfulBuild/artifact/tmp/en-US/html-single/index.html#idm140371751126480

Let me know, if there are any more details required to be added in this section. If it looks Ok i shall move it to on_qa

The doc text looks good. I also have one more additional command to be added in section "7.3.7.1. Enabling Metadata Caching" as a 4th point:

4. To increase the number of files that can be cached:
# gluster volume set <VOLNAME> network.inode-lru-limit <n>
where n by default is 16384, it can be increased if the number of active files in the volume is very high. Note that increasing this number increases the memory footprint of the brick processes.


Rajesh,
Can you please take a look at this, Do we need to provide a number in the command?

Hi Poornima,

I have updated the section accordingly.

http://ccs-jenkins.gsslab.brq.redhat.com:8080/job/doc-Red_Hat_Gluster_Storage-3.2-Administration_Guide-branch-bz-1374150-gluster_nfs_off_by_default/lastSuccessfulBuild/artifact/tmp/en-US/html-single/index.html#idm140484642738080

Rajesh,

Let me know if you have any further comments before I move this bug on qa.

Poornima: We need not provide a number for inode-lru-limit as of now. It should be derived from workload.

Bhavana: As discussed we should move the md-cache section under the performance improvement of RHGS section. Samba section should refer to the new performance section.

Based on my meeting with Rajesh and Poornima, the section is now moved to Chapter 13 Configuring Red Hat Gluster Storage for Enhancing Performance as Section 13.1 Directory Operations.

Minor edits were made to this section to generalize the statement wrt RHGS instead of just SMB.

A link to section 13.1 is provided in the SMB section "Sharing Volumes over SMB"

http://ccs-jenkins.gsslab.brq.redhat.com:8080/job/doc-Red_Hat_Gluster_Storage-3.2-Administration_Guide-branch-bz-1374150-gluster_nfs_off_by_default/lastSuccessfulBuild/artifact/tmp/en-US/html-single/index.html#sect-Directory_Operations

The Section is 13.7, (and not 3.1 as mentioned in comment 11)

http://ccs-jenkins.gsslab.brq.redhat.com:8080/job/doc-Red_Hat_Gluster_Storage-3.2-Administration_Guide-branch-master/lastSuccessfulBuild/artifact/tmp/en-US/html-single/index.html#sect-Directory_Operations

The content is fine as mentioned in the link but as per recent decision on providing a single cli option to perform all these md-cache related changes via one command https://bugzilla.redhat.com/show_bug.cgi?id=1417955 , we need to add the cli option as well to the performance section.

Moving this bug back to assigned as it needs some new changes to be added.

Based on the recent changes, the new cli option is now added to the "Directory Operations" section:

http://ccs-jenkins.gsslab.brq.redhat.com:8080/job/doc-Red_Hat_Gluster_Storage-3.2-Administration_Guide-branch-master/lastSuccessfulBuild/artifact/tmp/en-US/html-single/index.html#sect-Directory_Operations

Looks good.Moving the doc to verified.

RHGS 3.2.0 GA completed on 23 March 2017