Document URL: https://access.redhat.com/documentation/en-US/Red_Hat_Storage/3.1/html/Administration_Guide/index.html
Additional information: POC - Ravi
We need to give a simple sizing recommendation to the customer. The recommendation in comment 12 about killing the data brick, running workload, etc will be an overhead for the administrator. Also, from the description, This sizing recommendation may not be accurate as the rate of space consumption will not be constant.
*** Bug 1361117 has been marked as a duplicate of this bug. ***
Shekhar, Manoj, can you also take a look at "Tunin recommendations for arbitrated volumes" in this document to be extra-sure it's okay?
Much of the content in the tuning recommendations section is not specific to Arbiter volumes. Can we put recommendations specific to Arbiter volumes in this section and point to existing sections for general recommendations?
Shekhar, can you help Laura rewrite the tuning recommendations as requested in comment #36?
Laura, As per Manoj's suggestion we can point to https://access.redhat.com/documentation/en-US/Red_Hat_Storage/3.1/html/Administration_Guide/chap-Configuring_Red_Hat_Storage_for_Enhancing_Performance.html for general recommendation.
In respect to Arbiter volume, only 1st two points under tuning heading holds true i.e.
For dedicated arbiter nodes, use JBOD for arbiter bricks, and RAID-6 for data bricks.
For chained arbiter volumes, use the same RAID-6 drive for both data and arbiter bricks.
The rest all can be covered under general recommendation.
I have following comments
0. It will be good to output the volume configuration for arbitrated replicate volume as captured from gluster volume info output
The section 6.8.3 should have the following post volume creation example:
Volume Name: testvol
Volume ID: ed9fa4d5-37f1-49bb-83c3-925e90fab1bc
Snapshot Count: 0
Number of Bricks: 2 x (2 + 1) = 3
Brick3: server3:/bricks/brick (arbiter)
Brick3: server6:/bricks/brick (arbiter)
1. Just above the figure 6.7, I see that the configuration is mentioned as 6*(2+1)
Please make it changed to 6 x (2+1)
2. "Converting to an arbitrated volume" should be renamed as "Converting to an arbitrated replicate volume"
3. The second example under "Converting to an arbitrated replicate volume" can provide the volume configuration mathematically.
Say, for example - The current content goes like this :
"If you have an existing two-way distributed-replicated volume, you need a new brick for each sub-volume in order to convert it to an arbitrated distributed-replicated volume"
It would be more easy to understand when the statement goes like -
"If you have an existing two-way distributed-replicated volume (2 x 2), you need a new brick for each sub-volume in order to convert it to an arbitrated distributed-replicated volume (2 x (2+1))"
To make the above said modifications, I'm moving this bug to ASSIGNED
In the section 6.8.3, following the volume creation command, could you please add the following: "# gluster volume info testvol" followed by volume information.
I have missed to provide the command in comment44
Added after the example command, assume this is the one you meant.
Fixed section is here:
(In reply to Laura Bailey from comment #47)
> Added after the example command, assume this is the one you meant.
> Fixed section is here:
Yes, you got me right.
verified with the changes cited in the doc_link in comment47
Moving to CLOSED CURRENTRELEASE since RHGS 3.2 GA was yesterday. All documentation is available from https://access.redhat.com/documentation/en/red-hat-gluster-storage/.