toasters November 2018

toasters@lists.teaparty.net

3 participants
6 discussions

Creating a readonly user to monitor NetApp cluster via check_mk
by Sylvain Robitaille 20 Dec '18

20 Dec '18

Hello fellow toasters users ... We're in the process of deploying a pair of new CDOT filers and one of the things we'd like to setup with them is having check_mk (nagios and rrd, under the hood) monitor various attributes. We have a working instance at the moment using snmp to query the filers, but check_mk also provides a "plugin" that claims to work with the filer's web interface, so we'd like to examine that. To that end, I've setup a read-only user on the filer, that I intend will be able to access the web interface: security login create -vserver fc1-na -user-or-group-name nul_cmk \ -application http -authentication-method password -role readonly \ -comment "check_mk monitoring" The user shows up when I "show": fc1-na::> security login show -vserver fc1-na -user-or-group-name nul_cmk Vserver: fc1-na Second User/Group Authentication Acct Authentication Name Application Method Role Name Locked Method -------------- ----------- ------------- ---------------- ------ -------------- nul_cmk http password readonly no none The readonly role is unchanged from the OS default: fc1-na::> security login role show -vserver fc1-na -role readonly Role Command/ Access Vserver Name Directory Query Level ---------- ------------- --------- ----------------------------------- -------- fc1-na readonly DEFAULT readonly security none security login password all security login role show-user-capability all set all 5 entries were displayed. Trouble is that if I try to login to the web interface as that user, the browser stays "stuck" at the "Loading system details..." spinner widget. It's been there now for about two hours as I type this. Of course, check_mk's plugin logs a timeout while trying to access the interface. I fully suspect that I'm missing something, of course, but I'm afraid I haven't been able to find what that is. If someone out there has experience creating such a user for readonly access, a pointer to where I should be looking and/or documentation I should be reading would be hugely appreciated. (cli is preferred because it's easier to document what I've done that way, but if someone points me to resources in the web interface I'm sure I'll be able to find what I need at the cli.) -- ---------------------------------------------------------------------- Sylvain Robitaille syl(a)encs.concordia.ca Systems analyst / AITS Concordia University Faculty of Engineering and Computer Science Montreal, Quebec, Canada ----------------------------------------------------------------------

1 1

Deleting a AD domain from a cluster.
by Jeff Cleverley 27 Nov '18

27 Nov '18

Greetings, We have a 8.3.2P9 MetroCluster. We bought another company a couple of years ago and have had 2 domains associated with the cluster. The first domain is going to be shutoff in a couple of weeks. We were not too worried about it since we are active on the go-forward domain, and have usermaps with both domains. There is a trust relationship between the 2 domains. For any of you who have experience with a previous AD domain being decommissioned, I would appreciate any information about problems that occurred. We just moved some 7-mode data (old filers on new AD domain) to new SVMs on the cluster. All went well but a couple of quirks came up that we can't explain. We set up local user accounts just like the old ones. All worked fine. File system permissions/owners were good. We realized that we forgot to set up WINS servers. We set up the servers, and the local accounts now show up as SIDs instead of the username. We also had a process using the new AD domain that would not work. When they used the old AD domain\user, it worked. Once we put the WINS server addresses in, the old AD domain\user account failed. We had to switch back to the new AD domain\user account. This indicates our local user accounts and other things seem to be tied into our old AD domain. I opened a call with NetApp to try and get more data on what may break. Thanks, Jeff -- ** Please note my email address has changed to jeff.cleverley(a)broadcom.com Jeff Cleverley Manufacturing Engineer 4380 Ziegler Road Building 1, Dock 1 Fort Collins, Colorado 80525 970-288-4611

1 0

FLI \ Flex Array Foreign lun error
by Chris Hague 11 Nov '18

11 Nov '18

Hi Toasters, When connecting 7 LUN's from a 3rd party array to an AFF700 initiator ports, randomly one or 2 of the LUN's go offline with the following error.. These LUN's (foreign disks) are all coming from the same 3rd party array and it doesn't happen to them all; Message: scsiblade.import.lun.offline: LUN 80G/X+MCr/LQ in volume with MSID 2155921120 on Vserver svm_fc has been brought offline due to an I/O failure on the underlying foreign disk with serial number 00000000000000000000000000000000000000000000000000000000000000000000000000000000 0000000000000000000000000000000000000000000000000. Description: This message occurs when a LUN currently being imported is brought offline by Data ONTAP (R) because of an I/O failure on the foreign disk. The issue is either that data cannot be read from or written to the source (foreign LUN), or that there is an internal error. Action: Verify the connectivity to the foreign disk. After resolving any connectivity issues, delete the LUN import and create it again. Running an; storage array config show Still listed 7 LUN's in the above case, but we have seen that drop to 5 \ 6 before, even though the LUNs are still presented and zoned. Un-presenting and re-presenting does seem to kick start them back into life, but is not ideal. This seems to happen on different vendor 3rd Party arrays (we have seen it happen both on HUS and HP presented LUNs) Can anyone offer any advice on this? Many Thanks in advance, Chris. #Apologies for the stupidly long disclaimer! This email is being sent by a subsidiary of Arthur J. Gallagher Holdings (UK) Limited, part of the Arthur J. Gallagher & Co. global group of companies. For details of the registered office, company number and, where applicable, regulated status of our subsidiaries, please visit https://www.ajginternational.com/legal-regulatory-information/. We are the data controller of any personal information you provide to us or personal information that has been provided to us by a third party. We collect and process information about you in order to arrange insurance policies and to process claims. Your information is also used for business purposes such as fraud prevention and detection and financial management. This may involve sharing your information with third parties such as insurers, reinsurers, other brokers, claims handlers, loss adjusters, credit reference agencies, service providers, professional advisors, our regulators, police and government agencies or fraud prevention agencies. We may record telephone calls to help us monitor and improve the service we provide. For further information on how your information is used and your rights in relation to your information please see our privacy notice at https://www.ajginternational.com/Privacy-Policy/. If you are providing personal data of another individual to us, you must tell them you are providing their information to us and show them a copy of this notice. Where you are obtaining a non-consumer policy of (re)insurance, or cover for additional risks or renewal under an existing policy, you are required to make a fair presentation of the risk to a (re)insurer which discloses every material circumstance which you know or ought to know relating to the risk to be insured. A circumstance is material if it would influence the judgment of a prudent insurer in determining whether to provide insurance for the risk and, if so, on what terms. Disclosure must be reasonably clear and accessible to a prudent insurer and made in good faith. The aforementioned duty of disclosure is the applicable duty under the laws of England and Wales. If your policy is not subject to English law you are expected to disclose risk information in accordance with the requirements of the applicable law. In such circumstances we expect you will disclose risk information at least equal to the standard required under English law and where the applicable law requires you to disclose information over and above the level required under English law you will provide such information in accordance with that law. Where you are obtaining a consumer policy of insurance, you must read each question and answer honestly and fully and must take reasonable care to not make a misrepresentation. Failure to comply with the above disclosure requirements, as they apply to you, could mean that your policy of (re)insurance is void, its terms are materially altered or that (re)insurers are not liable to pay all or part of your claim(s). If you are in any doubt as to your obligations you should ask your usual contact. This e-mail and any attachments are CONFIDENTIAL and may contain legally privileged information. If you are not the intended recipient of this e-mail message, please telephone or e-mail us immediately, delete this message from your system and do not read, copy, distribute, disclose or otherwise use this e-mail message and any attachments. Although the above company has taken reasonable precautions to ensure this e-mail and any attachments are free of any virus or other defect that may affect your computer, it is the responsibility of the recipient to ensure that it is virus free and the above company does not accept any responsibility for any loss or damage arising in any way from its use.

1 0

Creating qtrees in flexgroups
by Allen, Mark A 10 Nov '18

10 Nov '18

Hello, Looking to have a flexgroup, fg1, which contains 2 subdirectories, bike and car, and create qtrees under each… Something like, /fg1/bike and /fg1/car. Then create the qtrees, /fg1/bike/qtree11 and /fg1/car/qtree22. This would enable a single flexgroup to be mounted on different clients (for example, mount -t nfs -o vers=3 IP:/fg1/bike /mnt/bikes) based on vehicle type. The alternative is to just use 2 flexgroups. Thanks! Mark

1 0

Volume re-home options
by Jeff Cleverley 07 Nov '18

07 Nov '18

Greetings, We are migrating off a 7-mode cluster to a cdot metrocluster. We have a number of requirements to keep the 7-mode names. Each 7-mode controller has ~90TB of used data and an average of 30 volumes with CIFS and/or NFS exports. We had planned on using a mix of xcp/robocopy to migrate to an existing SVM, rename the old cluster, create new cdot SVMs, then re-home the volumes to the new SVM. We tested this on a new cluster and it worked fine. The new cluster was not MC and the re-home won't work with MC. Is there a way to get the MC to assign the IP of the old 7-mode name to one of the SVMs something like an alias? Thanks, Jeff -- ** Please note my email address has changed to jeff.cleverley(a)broadcom.com Jeff Cleverley Manufacturing Engineer 4380 Ziegler Road Building 1, Dock 1 Fort Collins, Colorado 80525 970-288-4611

1 1

Qtrees in FlexGroups
by Allen, Mark A 06 Nov '18

06 Nov '18

Hello, I have a 4 node cluster with 2 aggregates per node and a FlexGroup volume that spans all nodes with 1 constituent member volume in each aggregate. I’m interested in determining where qtrees are located within the FlexGroup volume. Does a cli command exist that can show which aggregate and constituent member volume a qtree resides within? Thanks in advance! Mark

1 0

2024

2023

2022

2021

2020

2019

2018

2017

2016

2015

2014

2013

2012

2011

2010

2009

2008

2007

2006

2005

2004

2003

2002

2001

2000

1999

1998

1997

toasters November 2018