ESXi 8.0.3 fail creating datastore on iSCSI LUN on PetaSAN
Pages: 1 2

admin
2,969 Posts
April 10, 2025, 10:16 pmQuote from admin on April 10, 2025, 10:16 pmTried VMware-VMvisor-Installer-8.0U3-24022510.x86_64.iso and it is working.
Are you using iSCSI Software Adaptor or some hardware initiator ? do have iSCSI 1 & 2 on separate vswitch/vmkerenel ?
Can you try on different hardware ? maybe an ESXi hardware/driver issue ?
Tried VMware-VMvisor-Installer-8.0U3-24022510.x86_64.iso and it is working.
Are you using iSCSI Software Adaptor or some hardware initiator ? do have iSCSI 1 & 2 on separate vswitch/vmkerenel ?
Can you try on different hardware ? maybe an ESXi hardware/driver issue ?

Ste
138 Posts
April 11, 2025, 8:45 amQuote from Ste on April 11, 2025, 8:45 amI'm using iSCSI software adapter (vmhba64), two separate vmkernels, each of them connected to a different physical NIC, on the same virtual switch, exactly in the same setup VMWare suggested for the other ESX 6.7 cluster (which is working fine, by the way). Both vmkernel ports are present in the Network port binding tab, both active, in fact available path to the LUN are immediately and correctly displayed.

I have 3 brand new servers (PowerEdge C6620), but they all have the same hardware configuration, and they behave in tha same way, so is not a hardware component malfunction. Unfortunately I do not have a different new server to test.
I'm using iSCSI software adapter (vmhba64), two separate vmkernels, each of them connected to a different physical NIC, on the same virtual switch, exactly in the same setup VMWare suggested for the other ESX 6.7 cluster (which is working fine, by the way). Both vmkernel ports are present in the Network port binding tab, both active, in fact available path to the LUN are immediately and correctly displayed.

I have 3 brand new servers (PowerEdge C6620), but they all have the same hardware configuration, and they behave in tha same way, so is not a hardware component malfunction. Unfortunately I do not have a different new server to test.
Last edited on April 11, 2025, 8:47 am by Ste · #12

admin
2,969 Posts
April 12, 2025, 9:34 amQuote from admin on April 12, 2025, 9:34 amOur recommendations is that each iSCSI subnet has its own/separate nic/vswitch/vmkernel.
What feels strange in the screenshot you posted, is that both ESXi client paths seem to connect to the same lun path/subnet at 192.168.101.193. For correct MPIO, each client to lun path should be on different subnet. If you use Dynamic Discovery and enter the first ip, it should return with both target ips, else you should be able to enter both under Static Discovery.
Our setup looks like:

Our recommendations is that each iSCSI subnet has its own/separate nic/vswitch/vmkernel.
What feels strange in the screenshot you posted, is that both ESXi client paths seem to connect to the same lun path/subnet at 192.168.101.193. For correct MPIO, each client to lun path should be on different subnet. If you use Dynamic Discovery and enter the first ip, it should return with both target ips, else you should be able to enter both under Static Discovery.
Our setup looks like:

Last edited on April 12, 2025, 10:10 am by admin · #13

Ste
138 Posts
April 12, 2025, 2:10 pmQuote from Ste on April 12, 2025, 2:10 pmFinally I got it !!
Tired of being stuck in this situation, today i erased and reinstalled everything from the beginning, also applying the very latest patches for both the vCenter8 and ESXi hosts. The result is that now it it working correctly ! I will never be sure about which was the exact responsibility (i suspect a bugged ISO image), but i could not waste more time on this issue.
Thanks anyway for your time and support. Ste
PS: by the way... the Broadcom support is really useless ! After 10 days they're still trying to understand the problem...
Finally I got it !!
Tired of being stuck in this situation, today i erased and reinstalled everything from the beginning, also applying the very latest patches for both the vCenter8 and ESXi hosts. The result is that now it it working correctly ! I will never be sure about which was the exact responsibility (i suspect a bugged ISO image), but i could not waste more time on this issue.
Thanks anyway for your time and support. Ste
PS: by the way... the Broadcom support is really useless ! After 10 days they're still trying to understand the problem...

admin
2,969 Posts
April 12, 2025, 7:37 pmQuote from admin on April 12, 2025, 7:37 pmExcellent. Nice to hear 🙂
Excellent. Nice to hear 🙂
Pages: 1 2
ESXi 8.0.3 fail creating datastore on iSCSI LUN on PetaSAN
admin
2,969 Posts
Quote from admin on April 10, 2025, 10:16 pmTried VMware-VMvisor-Installer-8.0U3-24022510.x86_64.iso and it is working.
Are you using iSCSI Software Adaptor or some hardware initiator ? do have iSCSI 1 & 2 on separate vswitch/vmkerenel ?
Can you try on different hardware ? maybe an ESXi hardware/driver issue ?
Tried VMware-VMvisor-Installer-8.0U3-24022510.x86_64.iso and it is working.
Are you using iSCSI Software Adaptor or some hardware initiator ? do have iSCSI 1 & 2 on separate vswitch/vmkerenel ?
Can you try on different hardware ? maybe an ESXi hardware/driver issue ?
Ste
138 Posts
Quote from Ste on April 11, 2025, 8:45 amI'm using iSCSI software adapter (vmhba64), two separate vmkernels, each of them connected to a different physical NIC, on the same virtual switch, exactly in the same setup VMWare suggested for the other ESX 6.7 cluster (which is working fine, by the way). Both vmkernel ports are present in the Network port binding tab, both active, in fact available path to the LUN are immediately and correctly displayed.
I have 3 brand new servers (PowerEdge C6620), but they all have the same hardware configuration, and they behave in tha same way, so is not a hardware component malfunction. Unfortunately I do not have a different new server to test.
I'm using iSCSI software adapter (vmhba64), two separate vmkernels, each of them connected to a different physical NIC, on the same virtual switch, exactly in the same setup VMWare suggested for the other ESX 6.7 cluster (which is working fine, by the way). Both vmkernel ports are present in the Network port binding tab, both active, in fact available path to the LUN are immediately and correctly displayed.
I have 3 brand new servers (PowerEdge C6620), but they all have the same hardware configuration, and they behave in tha same way, so is not a hardware component malfunction. Unfortunately I do not have a different new server to test.
admin
2,969 Posts
Quote from admin on April 12, 2025, 9:34 amOur recommendations is that each iSCSI subnet has its own/separate nic/vswitch/vmkernel.
What feels strange in the screenshot you posted, is that both ESXi client paths seem to connect to the same lun path/subnet at 192.168.101.193. For correct MPIO, each client to lun path should be on different subnet. If you use Dynamic Discovery and enter the first ip, it should return with both target ips, else you should be able to enter both under Static Discovery.
Our setup looks like:
Our recommendations is that each iSCSI subnet has its own/separate nic/vswitch/vmkernel.
What feels strange in the screenshot you posted, is that both ESXi client paths seem to connect to the same lun path/subnet at 192.168.101.193. For correct MPIO, each client to lun path should be on different subnet. If you use Dynamic Discovery and enter the first ip, it should return with both target ips, else you should be able to enter both under Static Discovery.
Our setup looks like:
Ste
138 Posts
Quote from Ste on April 12, 2025, 2:10 pmFinally I got it !!
Tired of being stuck in this situation, today i erased and reinstalled everything from the beginning, also applying the very latest patches for both the vCenter8 and ESXi hosts. The result is that now it it working correctly ! I will never be sure about which was the exact responsibility (i suspect a bugged ISO image), but i could not waste more time on this issue.
Thanks anyway for your time and support. Ste
PS: by the way... the Broadcom support is really useless ! After 10 days they're still trying to understand the problem...
Finally I got it !!
Tired of being stuck in this situation, today i erased and reinstalled everything from the beginning, also applying the very latest patches for both the vCenter8 and ESXi hosts. The result is that now it it working correctly ! I will never be sure about which was the exact responsibility (i suspect a bugged ISO image), but i could not waste more time on this issue.
Thanks anyway for your time and support. Ste
PS: by the way... the Broadcom support is really useless ! After 10 days they're still trying to understand the problem...
admin
2,969 Posts
Quote from admin on April 12, 2025, 7:37 pmExcellent. Nice to hear 🙂
Excellent. Nice to hear 🙂