Re: [Nolug] linux and netapp

From: Jj Gmail <>
Date: Thu, 25 Jul 2013 13:07:13 -0500
Message-Id: <>

Figured it out. The Lun had a thin mirror on it. The mirrored portion was not large enough to accommodate the the space presented to the Linux server. That was why Linux showed it had space but the total Lun did not. Don't ask why it was set that way by the storage admin.
It's hard to know what is happening in other departments. I liked it better in the days when servers had its own set of physical disks. No wonder why as400 admin are always smiling.

On Jul 25, 2013, at 10:45 AM, Ron Johnson <> wrote:

> What happened to "report disk full and make the application programmer do the right thing"?
> On 07/25/2013 08:47 AM, Jj Gmail wrote:
>> All other sans I have worked with, emc, IBM fastt, the operating system would report disk full and kill the process. Is Netapp different or is there a configuration.
>> On Jul 24, 2013, at 7:38 PM, Scott Harney < <>> wrote:
>>> What jimmy says is accurate. If you entirely fill up a lun on netapp, data ontap (netapp os) will take the lun offline from the host to prevent data loss and/or corruption.
>>> On Friday, June 21, 2013, Jimmy Hess wrote:
>>> On 6/20/13, j j < <javascript:;>> wrote:
>>> > Anyone else have any experience with linux connected to a netapp
>>> san? I
>>> > have had 2 seperate occaisions where a data lun, with no os or
>>> > page partitions on it, filled up and hung the server. Trying to
>>> figure out
>>> Usually... if the disk space available for a LUN runs out, and writes
>>> can no longer be accepted, there are going to be problems regardless
>>> of what operating system or client hardware has the initiator using
>>> that LUN.....
>>> > if it is a CISCO UCS blade, Netapp, or combination of both issue. Don't
>>> > think it is red hat.
> --
> "Mundus vult decipi, ergo decipiatur"
> Petronius
> ___________________
> Nolug mailing list
Nolug mailing list
Received on 07/25/13

This archive was generated by hypermail 2.2.0 : 07/26/13 EDT