Issue with Netapp and Tomee8 Build Pack

[dakshina]

Dear Sir/Madam We use ONTAP Netapp. We are unable to write files to the Netapp volume with Tomee8 online build pack. The previous version of Tomee works fine. The issue arises with Tomee8. I am sorry if I misstated something. I am quite new to this technology and I don't know any more info. By the way, we use PCF (Pivotal Cloud Foundry) as well. Our only choice , at the moment, is to revert to older version of Tomee(7 I believe). Any help form you is appreciated.

Thank you

murthy

[AlexDawson]

Hi there!

Sorry to hear you're having troubles with this - can you share details about controller model, ONTAP version, connection method (NFS, iSCSI, FC, etc) and some information about what error is seen?

[NetApp_SR]

Are there error messages displayed? If so please share the message and when it is displayed.

[dakshina]

Dear Sir

We use NFSv3 and v4 volumes . Please see: https://code.cloudfoundry.org/nfs-volume-release/

There are no error messages. This is all the information I could get from the DevOps team. Any help is appreciated.

Thanking you

murthy

[NetApp_SR]

The link you sent has some troubleshooting for NFS mounts. Are the developers able to mount and unmount? From the storage side we can enable auditing to see what access is being requested and if it is successful.  Enable auditing and then ask the developers to attempt to access the mounts. The links below are for Clustered ONTAP. If you have 7 Mode let me know and I will look for other links.

How to enable auditing of NFS events on clustered Data ONTAP

https://kb.netapp.com/app/answers/answer_view/a_id/1031357/~/how-to-enable-auditing-of-nfs-events-on-clustered-data-ontap-

Auditing NAS events on SVMs

https://docs.netapp.com/ontap-9/index.jsp?topic=%2Fcom.netapp.doc.dot-cifs-nfs-audit%2FGUID-198A27B9-CC9B-457B-BD87-7842E5C8B262.html&cp=13_4_1

[dakshina]

Dear Sir Thank you for the response. The DevOps team is reluctant to work with us because Tomee is open source and they have some reservations about granting further access to their systems. We are basically caught between a rock and hard-place meaning some of our java jars (libraries) are incompatible with older versions of Tomee and these newer jars are necessary to get past the security scanners. I can pass on the information you have kindly provided to DevOps May I suggest you to try it by downloading and installing Tomee8 (https://tomee.apache.org/download-ng.html) on a test machine. All you have to do is write a file to the NFS mount. I can help you with it by setting up a servlet (please see https://www.roseindia.net/answers/viewqa/JSP-Servlet/9798-Writing-a-file-using-servlets.html). By the way, the DevOps tell me they are able to mount and unmount. It still works with older versions of Tomee. Thank you murthy

[dakshina]

Hi Alex I have gathered some additional information about the NetApp and PCF (Pivotal Cloud Foundry) Container. Here it is The problem seems to be in PCF container can write into the NetApp but after the file is written it is not readable. The permissions _rw_rw_r as compared to normal _rw_rw_rw. The controller model is NetApp ONTAP 9.6 The mystery deepens: I am informed that on one instance (call it A) of Tomee8, there is no issue. It is the second Tomee8 (call it B) that is giving raise to this issue. Both Tomee's are of same versions. Just to reiterate, when we downgrade the second Tomee8 (instance B) to Tomee7 everything is fine with NetApp. So it makes me believe that it is not with the code our application is based on. Your kind help is appreciated. Thank you murthy

[AlexDawson]

Hi there!

So it sounds like the UMASK (default file permissions) of Tomee8 is not correct. There is a new version of Tomee8 released recently (20 April) which fixes a bug relating to this - https://tomcat.apache.org/tomcat-8.5-doc/changelog.html

Can you verify if you are running 8.5.54 on both systems? and if not, try doing the upgrade and see if that helps.

[dakshina]

Hi Alex

I found the error is because of configuration of an environment variable in Linux server upon which the NetApp is mounted. I also got a hold of our NetApp provider and he couldn't tell me definitively if the error was because of application as he was looking at security trace. We never did the audit.  

While at it, I heard that some AWS customers, who don't mind the latency, are preferring to save their files in S3 as it is cheaper. Are you aware of that? Is there any way NetApp can make the pricing competitive with AWS?

Thank you

Murthy

[AlexDawson]

Hi There!

Great to hear it's all fixed. 

With regards to your question on AWS - We've got a number of solutions for using AWS.

The simplest one is our fabricpools for SSD aggregates - you can have a small SSD aggregate, configured to use S3 from AWS or another S3 provider for files which are not frequently accessed, or for previous versions of files from snapshots.

Other options include virtual NetApp systems in AWS to use as a snapmirror or snapvault target for backups, storage for Kubernetes and openshift, and many more. We have all this detailed on this page on our cloud site - https://cloud.netapp.com/aws-partners

If you have any questions about the options on that page, just click the little blue icon in the bottom right of the page and someone will get back to you in real time.

Thanks!