So as you can see I am using a ReadyNas RR4312X. I have been having problems with it where I have network drops when I have snapshots on. It seems to be worse when the snapshots are being deleted along side new ones being created.
You can take a look at the image I added to see the problem I encountered when testing the unit.
Looking at your charts, I'm seeing peak network speeds of ~35KB/s with snapshots disabled, and ~10 MB/s with snapshots enabled. Both are low (but ignoring the drops/spikes you are actually seeing much faster speeds in the bottom graph).
What firmware are you running?
How are you putting a load on the network? Is that program running over gigabit ethernet or is it running over WiFi?
Have you looked in the log zip file for disk errors or btrfs errors?
I have been saving video recordings (constant streams) to the NAS unit and I had virtual machines saved on it as well. I needed to move them off so I could use them.
I have also seen a spike in the temperature when the network drops. Mainly the CPU.
This should be running over the 10G port with a CAT 6 cable.
I did check the logs to see if the disks had any errors or warnings but everything seemed fine as there were no alerts.
I did check the logs to see if the disks had any errors or warnings but everything seemed fine as there were no alerts.
The alerts aren't as helpful as they should be - IMO the thresholds are much too high. So maybe look in the log zip file (disk_info.log will give you the SMART stats, you can also look for disk errors and btrfs errors in system.log and kernel.log ).
I recently found a failed disk in my own NAS that didn't have any SMART issues. I found UNCs when I ran smartctl -x from ssh. So perhaps do that as well.
Just upgraded the to the stable version of 6.10.1.
It is set to Raid 5.
The switch is not a10G im currently using so it wont really help.
I spoke to the person who tried to fix this before. He said he looked at the log file for a while to try find any disk errors and came up with nothing. He still suspects it to be the disks. Reason for this is that some of the drives had been stalling before in another server. We still cannot prove anything is wrongh with the drives though. Any ideas on how to could test them? We already used the seagate tool to run tests of them and all came back fine.
these drives are seagate survaliance drives. We are concidering getting drives that are more suited for the NAS.
We still cannot prove anything is wrong with the drives though. Any ideas on how to could test them?
Did you try running smartctl - x as I suggested above? Not sure if it will give you anything on your particular drives, but it is easy to do.
You could also capture the log zip file shortly after a network drop, and then submit the zip file to Netgear mods here for their analysis. Do that with a Private Message (PM) - click on the envelope icon in the upper right of the forum. Include a link to the zip file in the message (google drive, dropbox, etc). Don't post a link here.
The switch is not a10G im currently using so it wont really help.
I was trying to get some sense of your network path, as it would give us some idea of what the network performance should be.
If the switch is managed, then perhaps look at the ethernet stats for the NAS switch port. Also look at the ethernet stats on the NAS (in network_settings.log).
Another simple thing to try is change the connection to use the 1 gigaport port on the NAS.
Though I am thinking this is more likely disk-related.
