I have a Gig local network. All devices are routed through gig connections. There is a dumb switch in the middle. My server a cache pool with 2 SSDs totaling 378gb, with a minimum of ~90 used for appdata and docker.

 

For as long as I can remember using the cache drives has ensured transfer speeds as expected. I don't do many transfers from my Windows 11 box to the server, maybe once per week on average. Today I transferred some files over to a cached share, each being between 30-50mb. Normally files of this size and type wouldn't even trigger the Windows Explorer transfer dialog box as they'd be near instant. This time it did and it caught my attention. I noted all transferred seemed to cap out ~12 MB/s. I then tried copying some files from my array to the Windows machine, and those transferred peaked ~11 MB/s. I then transferred some other small files to a different share and saw speeds from 8-10 MB/s. 

 

I have Netdata container installed, and some other dashboard types as well. But to be honest, I haven't put the time in yet to learn to interpret Netdata's output. I'm really at a loss troubleshooting. Could it be a problem with the an SSD in the cache pool? Something in the switch I should be checking? Are there numbers I can grab from Netdata or another monitor to help identify the bottleneck? The only thing I have changed in the config this week was creating a custom network for Docker and moving all my containers over to it. Can that have something to do with it?

 

Any guidance on what to troubleshoot would be greatly appreciated. I'm attaching my diagnostics .zip. Please let me know if any other logs are needed and I will oblige.

mootower-diagnostics-20220507-2047.zip

Edited May 8, 20224 yr by VelcroBP
typo in cache pool size

per a reddit thread I ran iperf with my unRaid as server and my windows box as client. Here are the results:

 

Unraid dashboard indicates it's registered full gig duplex, so does this mean my switch might be at fault? What to check from here?

Hmm, seems like my Windows 11 box is the culprit. Wonder what's what here but it looks like it's not an unRaid issue after all...