ReadyNAS 2120 V2 hangs alost every month and LUN connection problem

[Alein]

Hi , I have some problem with this hardware.

1.Sometimes my ESXi reboot, and LUNs are not able to reconnect.  I see that, when I will try reboot NAS, I am warn that there are still active LUNs.

This is big problem,  I have to reboot NAS. Is there a way to see active LUNs and disconnect some of them? This option should be in menu, but it is not. Is it possible from SSH?

2.Almost every month NAS i hanging on 6.5 firmware. This is crazy. Any one of you have simmilar problem?

[Alein · Accepted solution]

It seems that iSCSI is useless on this devices.
Since I had removed whole iSCSI configuration, my NAS become much more stable, even with snapshots and NFS datastore for ESXi.

So my recomendation, Disable iSCSI.

[mdgm-ntgr]

You should investigate why your ESXi server is rebooting. You should always reboot it safely.

The RN2120 is a cost effective rackmount device designed for light use business cases.

How many LUNs do you have and how many VMs are they for?

Are you using thin or thick LUNs?

Are you using snapshots?

What about sync writes.

You should be using thick LUNs, with no snapshots and sync writes disabled.

You should also turn off VAAI in ESXi.

If the wrong settings were set you may need to backup your data, delete the LUN and create a fresh LUN.

[Alein]

Thank you for fast response.

I am loosing connection with hypervisor, the main reason is that it is on USB drive, and sometimes under high load it is happening. 

But this is not the point here.

I am turning off all VMs, and forcely reboot server.

After that some LUNs are not reconnected.

VM use iscsi as storage(backup,data). OSes are only 4.

Currently I have 3 LUNs 

All are THIN, I use snapshots (daily),sync writes ON, VAAI ON.

So basically it is oposite to what you wrote 🙂

Is there any reason why I should not use VAAI,snapshots,thin,and daily snapshots?

[mdgm-ntgr]

Using thin LUNs will impact performance as the LUN has to be grown over time and may become fragemneted. For ESXi you want to configure things to optimise performance and thick LUNs are the way to go.

Snapshots use CoW at the point in time they are taken. With something that has a huge level of writes to it in place you're going to get a lot of fragmentation. A VM is a classic example of this.

Likewise bit-rot protection should be disabled (we link this to enabling/disabling) CoW.

CoW (Copy on Write) means that when changes are made the existing data is not overwritten, but instead written to free space. So you will get a huge level of fragmentation using it with VMs.

CoW is enabled per LUN/share, and the way to undo it would be to backup your data, delete the LUN, create a LUN with CoW disabled and restore your data from backup.

Sync writes are good for verifying writes are synced properly, but this does impact performance which is a problem for VMs. With VMs you want to configure things to optimise performance.

We don't support VAAI and we mention that in this article. Trying to use a feature that is not supported is going to lead to issues.

[Alein]

Ok, some of your points makes sense.
But this is not answer on my question.
How can I list active LUNs and disconnect some of them?

[Alein]

BTW.

Are you sure that snapshots are turning on CoW? I do not use BitRot Protection.

[StephenB]

---

@Alein wrote:

Are you sure that snapshots are turning on CoW?

---

Yes, and they have to. Snapshots can't work without out.

[mdgm-ntgr]

You would use VMWare's software for safely stopping using the iSCSI storage so the NAS can reboot safely.

[Alein]

I have 3 such NASes.

All of them have got this same problems.

I am removing ISCSI from one of them to see the result.

At this point even 1 VM on SCSI can freeze NAS. It is not acceptable.

I do not use VAAI, CoW, I use asynchronous tranfer mode, All of them are working with NIC aggregation LACP with netgear switch and jumbo frames 9000.

I use snapshot ;daily. I just need them.

So at this point this is no acceptable that this NAS is freezing.

Can I get support , Can I send logs to someone? This must be an OS6 issue.

And there is still problem with locked LUNs, after ESXi reboots.

[mdgm-ntgr]

Well redo your setup but this time do it per the recommendations I gave above and see if you still have issues.

[Alein]

Well,

I made upgrade to OS 6.6 and my NAS died 😞 There is no ping, he is not trying to get IP, Reinstal OS is not helping.RAIDAR is not detecting it.

Any idea what can I do? I can not downgrade as this is new Debian and there is such information.

[StephenB]

You might need to use netgear (paid) support.

[Alein]

It is really buggy FW.

After upgrade my LAG was reconfigured to default settings.

Now I have issue with NFS41

vmware is telling me that it is not supported 😞

WARNING: NFS41: NFS41ExidPathProcess:2095: Server doesn't support the NFS 4.1 SunRPC program

[Alein · Accepted solution]

It seems that iSCSI is useless on this devices.
Since I had removed whole iSCSI configuration, my NAS become much more stable, even with snapshots and NFS datastore for ESXi.

So my recomendation, Disable iSCSI.