Bug 1412117

Summary: [RHV-RHGS]: Parallel IOs from two sub directories of VM is not working.
Product: [Red Hat Storage] Red Hat Gluster Storage Reporter: Byreddy <bsrirama>
Component: replicateAssignee: Krutika Dhananjay <kdhananj>
Status: CLOSED NOTABUG QA Contact: SATHEESARAN <sasundar>
Severity: high Docs Contact:
Priority: unspecified    
Version: rhgs-3.2CC: bsrirama, pkarampu, rcyriac, rhs-bugs, sasundar, storage-qa-internal
Target Milestone: ---   
Target Release: ---   
Hardware: x86_64   
OS: Linux   
Whiteboard:
Fixed In Version: Doc Type: If docs needed, set a value
Doc Text:
Story Points: ---
Clone Of: Environment:
Last Closed: 2017-02-15 10:53:03 UTC Type: Bug
Regression: --- Mount Type: ---
Documentation: --- CRM:
Verified Versions: Category: ---
oVirt Team: --- RHEL 7.3 requirements from Atomic Host:
Cloudforms Team: --- Target Upstream Version:
Embargoed:
Bug Depends On:    
Bug Blocks: 1277939    

Description Byreddy 2017-01-11 09:59:25 UTC
Description of problem:
=======================
Parallel IOs from two sub directories of VM is not working as per the expectation, IO from one sub DIR is going for pause state when IO was happening on other sub directory.


Version-Release number of selected component (if applicable):
=============================================================
glusterfs-3.8.4-11


How reproducible:
=================
Always


Steps to Reproduce:
===================
1.Have RHV-RHGS SETUP with 3 rhgs nodes and 2 clients (hosts)
2.create a 2 *3 volume 
3.create a Application VM using the vm storage created in step-2
4. create two sub directories (say mkdir /home/terminal{1..2} in the app VM
5. open two terminals of VM.
6. start dd io on terminal1 directory and kernel untar from other sub DIR parallel // you will see IOs going for paused state for long time. 

Actual results:
===============
Parallel IOs from two diff directories of VM is not working.


Expected results:
=================
Parallel IOs support should be there

Additional info:
================
I will provide the required  logs

Comment 2 SATHEESARAN 2017-01-11 10:05:00 UTC
This may affect grafton too, if there are more than one applications running on the application VM and consuming the storage, it may lead to one app to starve.

Comment 3 Krutika Dhananjay 2017-01-13 04:56:40 UTC
As discussed on Wednesday, please provide volume-profile output of the run the next time you try to recreate it, for further investigation.

Comment 4 Byreddy 2017-01-13 05:24:29 UTC
(In reply to Krutika Dhananjay from comment #3)
> As discussed on Wednesday, please provide volume-profile output of the run
> the next time you try to recreate it, for further investigation.

Right now setup is busy with diff issue...i will provide the cmd output.

You can easily reproduce this issue.

Comment 5 SATHEESARAN 2017-01-13 11:04:46 UTC
Byreddy could hit this issue in his setup consistently, but with the same workload, I couldn't hit this issue in my setup ( QEMU/KVM RHGS setup ).

The only difference being that Byreddy has RHGS machines are Virtual machines and in my setup I have RHGS machines as physical machines and volumes are mounted back on the same hypervisor

We are testing it out to eliminate differences in the setups and could arrive at the real problem

Need more time to investigate whether this is a real issue or a specific setup issue

Comment 8 Pranith Kumar K 2017-02-10 07:22:50 UTC
Sas,
      For 3.3.0 planning, what is your recommendation about this bug?

Comment 9 SATHEESARAN 2017-02-10 11:47:53 UTC
(In reply to Pranith Kumar K from comment #8)
> Sas,
>       For 3.3.0 planning, what is your recommendation about this bug?

Pranith,

As I heard last from Byreddy that this issue is no longer occurring in his setup.

@Byreddy, could you please confirm ?
Please close this bug, if its no longer reproducible or occurring in the setup.

Comment 10 Byreddy 2017-02-15 10:53:03 UTC
(In reply to SATHEESARAN from comment #9)
> (In reply to Pranith Kumar K from comment #8)
> > Sas,
> >       For 3.3.0 planning, what is your recommendation about this bug?
> 
> Pranith,
> 
> As I heard last from Byreddy that this issue is no longer occurring in his
> setup.
> 
> @Byreddy, could you please confirm ?
> Please close this bug, if its no longer reproducible or occurring in the
> setup.

When i tried once again this issue  on good configuration VMs, i didn't see this issue so closing as NOTABUG