I am running two almost identical ReadyNAS NV+ units in a network. One has terrible performance problems, the other is fine. The specs are as follows
Good unit: firmware v4.1.7, 1GB RAM, 4 x WD20EARS-07MVWB0 disks, ssh enabled Bad unit: firmware v4.1.10, 1GB RAM, 3 x WD10EADS-00M2B0 disks (96 GB (5%) of 1838 GB used), ssh enabled. Factory reset recently performed.
The problem: copy operations to the NAS over CIFS is slow and repeatedly stalls. In these moments the window on the computer that is performing the copy is not redrawn. I monitored the NAS and saw that the CPU is in the iowait state >90% in these moments.
See the diagrams showing a copy operation to both of the NASes from the same source, using the same computer. Note the horizontal parts of the graph when copy was stalled.
I also tried mounting the source over NFS on both units and copy the same files. Copying to the bad unit took 5 times longer than to the other one.
I replaced the memory module in the bad unit but that did not help (used a 512MB module instead of 1GB). The strangest thing: I then put the RAM module from the good unit into the bad one. It did not even boot with that module. I put the RAM back into the good unit and it works fine.
Now where should I look for the cause of this problem?
Thanks for the suggestion. I replaced the disks with a single small spare HDD. Tests showed performance was on par with the other, good unit. Seems the NV+ doesn't like the WD10EADS disks. I'll order three WD20EFRX HDD's and replace them.
