ReadyNAS 2100 - Deleting a large file on the NFS

[scgsg]

Regarding the issue detailed:

http://kb.vmware.com/selfservice/microsites/search.do?language=en_US&cmd=displayKC&externalId=1035332

Basically the issue is that when a large file is deleted/removed from a NFS datastore on the NAS, the datastore becomes inactive and we get an error 'No such file or directory' (eventually). When the datastore becomes active again, we find that the file is deleted. According to the KB the workaround is to increase the timeout but before having to do that (given that the host has to be rebooted), is it possible to resolve the issue on the NAS?

[mdgm-ntgr]

If there is a change possible at the NAS end it would be recommended to power off the VM first. A change could lead to the NFS service needing to be restarted.

[scgsg]

The datastore does not have any active VMs as its more of a backup. The issue is cloning to the NAS or Deleting files from the NAS. The short of it, I'm not worried about rebooting the NAS. I just need to know if there's a way to configure it?

[scgsg]

Is there no known way to alter the response time to the NFS Delete command?

[mdgm-ntgr]

On the NAS side you could try using async.

The article you linked to suggests a fix changing a setting on the ESXi host.

[scgsg]

Can you point me to the appropriate documents or commands or information about this?

As i already said, I'd rather not make the change to the Host as it requires a reboot. Additionally, its not a 'fix' its a workaround.

[mdgm-ntgr]

Go to Share > Share Listing > Share name > NFS.

Under Advanced NFS options uncheck Enable Sync Mode and click Apply.

[scgsg]

That option is not available.

[OlivierB]

Hello,

I permit myself to respond that you will never had a good response from Netgear on this problem.

They know it : it exists from several years now, there are post in 2009, 2010 of the same problem.

New versions of OS have never change anything.

When you use esx with these materials in Iscsi or in NFS (i tried the 2), you'll always have problem.

Just now i deleted a 2 TB vmdk on one of my netgear nfs datastore, i stopped the only one vm before, the checkbox is uncheked and i have the probleme, 

Waiting during 10 to 20 minutes before the file is deleted (you can see in WebUI the free space that is growing), i said myself "i'll google to see if something has changed previous year", find your post i knew that nothing changed and i think nothing will nether change because they don't mind.

I don't know if it's only some models or alls but i have 6 readynas 2400 and it's the same for all.

good luck.

[kohdee]

This is not a specific ReadyNAS NFS issue. This is a pretty common NFS issue across many platforms. A Google search turned up a few other NAS providers who do go through this same thing. It'll be unlikely to be resolved in RAIDiator 4.2 but if you can show me some where other companies are resolving such problems, we can probably see about adding something in OS 6.

[kohdee]

Deleting a huge file over iSCSI should not be a big deal for any ReadyNAS because the ReadyNAS has no idea what is going on inside of your LUN and that file system. 

[OlivierB]

For my experience it is specific to readynas.

I tested it on compellent and qnap and had no problems in nfs.

It is the only reason for wich i leave my readynas to qnap for small sans.

I'm not an OS builder, i'm a customer that simply want materials that function correctly and a constructor that correct errors.

[OlivierB]

for iscsi :

I remember that i had the same problems with iscsi, but i won't waste more time by now.

Did you try deleting a big iscsi lun when the san is already in use with other luns in iscsi ?

I would be pleased to have the response.