Not creating nvme namepsace

[peleg2312]

I deployed trident nvme/tcp with San protocol, I am using the parallel feature and with 100 workers in the csi provisioner container, when running and creating 200 pvc for testing I see it take 6 min and for some reason there are some volumes created without nvme namespace, there are 160 nvme namespace and 200 volumes but for some reason the pvc in bound witch is weird? And the netapp take a lot of time between creating each volume and the nvme namespace,

I did try with 20 workers and it worked fine but I want it faster and better because my cluster is very heavy on this so it can become a problem

Someone know about It???

[peleg2312]

I am using netapp 9.15 and trident 25.10

[clintonk]

Hi, @peleg2312. Thanks for trying the concurrent tech preview! I agree that having 200 bound PVCs but only 160 namespaces doesn't sound right. Did you see any errors in the Trident controller logs when that happened? Would you be willing to share relevant log snippets here? FWIW, 20 workers from the provisioning sidecar is likely to drive a single ONTAP SVM near its API limit, and it's already 20x better than having everything serialized in Trident.