6.2.7 - 802.3ad Link Aggregation Issue


koz97

Recommended Posts

Hi Guys,

 

I have just currently installed a managed switch on my home network. Since doing this I have created a bond (2x 1Gbps NICs) on my UNRAID server which now gives me 2Gbps of bandwidth. The bonding mode I have chosen is 802.3ad since my switch supports LACP. 

 

On the switch side of things I have created a trunk, which includes 2 x 1Gbps ports being teamed together, which is also using the LACP protocol.

 

Once completing these two steps, I conducted a test where I copied two large ISO files to my UNRAID server at the same time. The two PCs from which I copied these two files from are utilising 1Gbps NIC cards. In theory, I should be able to copy these ISO files at the same time from each computer at around ~100MB/s. Although, once putting this into practice I was only getting around ~50MB/s on each computer, giving me a total of 1Gbps of bandwidth to the UNRAID server. 

 

Is there anything I might be doing wrong? Or is there possibly another bottleneck on my server which is slowing it down to 1Gbps? 

 

Any help will be much appreciated.

 

Thanks

Link to comment

First, does both port LACP dynamic status indicate success on switch side? Or simple if you found traffic (LED) at both port.

 

I suggest you test by 2 PC and each read access 2 file from different unraid disk to verify LACP setup, if found could utilize the 2Gbps, then bonding is functional.

 

There are several post about LACP setup not success.

Edited by Benson
Link to comment
On 8/31/2019 at 2:17 AM, Benson said:

First, does both port LACP dynamic status indicate success on switch side? Or simple if you found traffic (LED) at both port.

 

I suggest you test by 2 PC and each read access 2 file from different unraid disk to verify LACP setup, if found could utilize the 2Gbps, then bonding is functional.

 

There are several post about LACP setup not success.

Yes, whilst transferring data to my UNRAID server, both ports on the switch that are apart of the trunk have activity fllickering on their LEDs. 

 

I'm quite certain the trunk configuration is correct on the switch side of things as I currently have another trunk running into another server of mine running Windows Server 2012 R2.

 

Is there anything I might be missing on the UNRAID side of things?   

Link to comment

So the LACP should work as expected.

 

To eliminate the write performance limit by storage (unraid), so I would suggest you perform read test instead of write test. If read test (each PC read a differeng file ) also ceiling at 1Gbps then it should be a bug.

 

You should haven't setting wrong or missing.

Edited by Benson
Link to comment

Just ran a test where I copied two ISOs from two different computers at the same time to a share which was located on my cache drive.

 

Both files were copying at ~100MB/s, thus achieving 2Gpbs bandwidth to the UNRAID server. 

 

So the bottleneck in my first test was my two array HDDs. Is this because there is a parity drive? 

Link to comment

Due to the nature of parity, LACP and spinnder disk limitation, there are impossible reach 2Gbps write speed, this not only because parity disk performance, it relate all of array disk.

 

If no parity, in simple say, you could got 2 100MB/s write stream to 2 disk.

Edited by Benson
Link to comment

Join the conversation

You can post now and register later. If you have an account, sign in now to post with your account.
Note: Your post will require moderator approval before it will be visible.

Guest
Reply to this topic...

×   Pasted as rich text.   Restore formatting

  Only 75 emoji are allowed.

×   Your link has been automatically embedded.   Display as a link instead

×   Your previous content has been restored.   Clear editor

×   You cannot paste images directly. Upload or insert images from URL.