Ask the Experts: EMC Isilon technical content and documentation

I found the isilon blog very useful. Moreover I felt SmartConnect, SmartPools, SyncIQ and protocols is what Isilon is all about. We really need to concentrate on these things before implementing a new Isilon at any customer site. I noticed onething with the authentication providers that when we are joining the cluster to a Windows server 2008 domain controller - things are working fine, but when it comes to joining the cluster to a Windows server 2012 R2 domain cluster in a "Primary / Secondary" domain controller environment - the SMB shares do not map automatically with any user credentials.

Second thing, In a total cluster reboot operation - we must do a NTP sync post reboot. I felt if each node is independent of each other then what is the point of making it a cluster ?

I may not have framed the questions properly but I hope someone understood them.

Hi Kirsten,

Thanks for your response. I would like to see an improvement in transparency in the part of release-pulls.

The NEW! Tags were a good idea, but they should be removed when the document is updated. So you only see the changes from version to version.

The Hub I use the most is Isilon Uptime Info Hub. It provides me the necessary information about target code, firmware, and links to the most important documents (Release Notes, OneFS Patches etc.). It mostly replaced my personal link-collection.

You didn't ask for it, but the most useless one in my opinion is the SyncIQ Info Hub. It is mostly merchandise for Superna and some alibi-Infos about SyncIQ.

Regards

-- Steffen

> So, what commands would you like to know how to use?

Andre,

that's a bit of the hen and the egg problem...

There are lots of isi_* commands  in /usr/bin, some which I might like to use

once I read the man page. Which many of the commands are lacking...

Like  isi_stf_test stat

Very helpful when we needed to figure out which snapshots contained gazillion of (small) files!

And still wondering wether isi_stf_test rmlin can be used to remove individual files from snapshots.

This kind of stuff.

> Or what information or tasks do you want to perform from the command line and are not sure which commands or options are available?

Lots of things, but  this ends up in feature or enhancement requests

(besides plenty of "cool" home-made scripts)

Few examples:

- isi statistics client   => like to order by throughput (In + Out MBs/)

- isi statistics client   => like to order by  "utilisation rate", i.e. product of  Ops[1/s] * TimeAvg[us]

- isi statistics query --top  => should have capability to order by any column, too, in particular for top-style

- isi statistics history  => need upfront list of history-tracking stats, similar to isi statistics list stats

- isi job status/events/statistics  =>  like to have a simple query for job status AND running time, one line per job (current + finished)

- any command that takes the --nodes option => be able to specify node pool or tier name as the node list

- being able to browse the /ifs tree to see what exports/snapshots/quotas/filepools/... are configured on directories,

   i.e. a filesystem-centric, unified view of all path-based configurations.

- Smartpools job reports should show not only files and bytes present per file pool,

  but also how many files and bytes have been actively restriped by the last job run (per file pool).

Hope that helps... you also might share what's on your mind to get other focussed feedback.

-- Peter

Hi Kirsten and team

documention is essential and a multi-level challenge.  References, concepts, training, issue/fix alerting -- all different things. And we have witnessed a wealth of new communication channels from Isilon in the past year or so, absolutely amazing and big kudos to you all!

Could you share a bit of background on how the team -- or multiple teams? -- work(s)?

It has appeared to me that establishing the info hubs in the forum somehow bypasses the more corporate-style information channels for Isilon:

https://support.emc.com/products/15209_Isilon-OneFS

https://support.emc.com/search/?text=*&product_id=15204&sortBy=DREDATE&sortOrder=descending

Both of which pages provide versatile querys, are linked to the e-mail notifications etc.

However these pages have been missing a bit of loving care recently...

Only three White Papers listed within the past 12 months? -- I remember Chris

had to double-check to make sure his paper gets listed.  And the references

to the community forum are stuck with a link to a LCD question from April 2014.

So, what's going on with these pages... did you give up on them and started

the info hubs instead, in a Wiki style?    It's a lot a manual work to get things

correctly cross-referenced, have useful news tracking etc. -- you have found it out...

Speaking of a Wiki, the hubs are editable by everyone -- is this meant to encourage

all of us to participate?

The Isilon forum blog has great stuff, and providing the opportunity for

discussion makes this medium superior to static KB articles, Tech Notes,

and White Papers. Actually there's the Isilon Knowledge Blog, too. Bit difficult

to keep track of all the goodies, in particular when searching back for something.

Could these channels be streamlined somehow -- but then again, how may teams

are at work here... ?

Btw, logging the recent changes would be great for KB articles and

all docs that occasionally get re-published,

just like with the Current Software Releases paper.

In the  ETA and ESA info hubs, could mark the items' status as

Open/Resolved alongside with New?

What exactly is the purpose of what appears to be the front landing page

for the Isilon forum?

Isilon

It's quite a mess, and might distract people from entering and finding

the community with real clients & Isilon experts interacting.

As for actual documention content:

In fact and authentification and network have something in common, they need to

be planned and set up with teams outside storage people. The Isilon authentification

White Papers can hardly be consumed by AD/LDAP admins, just to specific

on Isilon CLI examples, which doesn't buy them much.

A higher level description and graphics of concepts, mechanisms

and scenarios that can be shared by Isilon *and* AD/LDAP admins can

foster communication and avoid misunderstandings.

One more thing -- the usual "it depends" answers in the forum,

when it comes to performance prediction for NAS traffic as well as

the OneFS job engine, and their mutual influence.

Having a couple of rich scenarios, with quantitative stats from

network to disks, and a couple of variations per scenario, can help

making educated predictions.

The release notes for drive support packages just list the

new firmware version numbers for affected drive types -- I'd like

to learn more about the actual improvements in order to make

an informed decision on the update.

Biggest issue by far for me is keeping track of OneFS software

releases, patches and know issues.

I like the new format of unified family and MR releases. Btw,

with a new minor family like 7.2.1, what are the initial "fixed issues",

fixed since 7.2.0 or some specific 7.2.x?

Now, rolling the drums for (quote form 7.1.1.0-5 MR notes):

Target Code known issues

The issues in the following table are known to affect OneFS 7.1.1.4 (Target Code). In addition, unless otherwise noted, issues that are resolved in OneFS 7.1.1.5 and later are known to affect OneFS 7.1.1.4. For a complete list of issues that are known to affect OneFS 7.1.1.4, you should also review the OneFS 7.1.1.5 and later resolved issues topics.

Frankly, I can't get around this. Where is something "otherwise noted"? What kind of Big Data analysis am I supposed to perform by hand here, in order to find out which OneFS release will be fit for my environment? I would need to cross-reference the notes with the available patches, too, of course.

Suggestion: Please (please) publish a table (Excel, csv) organized by issue id and code family,

listing the issue status with the family, e.g.:

Issue 99999  | OneFS 7.1.1 | fixed in patch 5555 for 7.1.1.3, fixed in 7.1.1.4

Issue 99999  | OneFS 7.2.0 | unresolved

...

...

That will, perhaps with some Excel macros or other scripting,

help make decisions on code versions to run under specific considerations.

Like "should I stay with some older MR plus patches, or move on to the newest family"?

I'm looking forward to the promised patches info hub, however I hope I made the point

that cross-referencing the available patches and releases by issue ids

(in and by the computer) will be highly beneficial.

Thanks for opening this ATE on Isilon documentation!

Cheers

-- Peter

Thanks Peter. Those are very helpful suggestions.

Chris Adiletta, who develops the Cool Commands segment of the ClusterTalk podcast, shared his methodology with me about how he picks what commands to include. He thinks about what a sys admin might want to accomplish based on these general categories:

1. What's going on with my cluster?
2. How do I get specific information?
3. How do I perform a specific task?

I could see the video series following this model and us concentrating a few episode in each of these categories. This is a new topic of research for me, so hearing from real users about their needs and fitting them into these categories will help me develop a more coherent plan that I hope will result in useful and interesting videos.

Hi Steffen,

Just wanted to let you know that we're in the process of scrubbing "NEW!" tags that are older than 3 weeks from our Info Hubs. Hopefully this will clean up some clutter. And thank you for the feedback about the Uptime and SyncIQ Info Hub. It's great to learn how everyone uses these Info Hubs, and which ones they prefer over others.

Thanks again!

Kirsten

One very cool thing built into  isi statistics client ...

is that --totalby can take multiple field name, comma-separated.

This provides very flexible choices for breaking out by whatever combination you want to see:

isi statistics client --totalby username,node

isi statistics client --totalby remotename,class

isi statistics client --totalby username,remotename,localname

etc...

Hi tazatemc,

Glad to hear that the Isilon blogs are useful! In regards to your technical questions about Windows 2012 authentication and reboot question, I suggest creating separate discussion threads on the Isilon community. Because this forum is dedicated to content discussion, it would give your technical questions more visibility among our community members. I'm also doing some research and can get back to you later.

Hope this helps you get answers to your questions!

-Kirsten

Thanks, Peter! These are helpful insights for us. I'm glad to address some specific points, and will do my best to answer your questions about the multiple teams that are involved in Isilon content creation and publication throughout this thread.

documention is essential and a multi-level challenge.  References, concepts, training, issue/fix alerting -- all different things. And we have witnessed a wealth of new communication channels from Isilon in the past year or so, absolutely amazing and big kudos to you all!

Thank you! We want to make sure that we're taking advantage of new communication channels that our customers want to use to receive information. However, the multitude of publication channels does lead to the challenge of keeping track of everything, as you mention.

It has appeared to me that establishing the info hubs in the forum somehow bypasses the more corporate-style information channels for Isilon: https://support.emc.com/products/15209_Isilon-OneFS https://support.emc.com/search/?text=*&product_id=15204&sortBy=DREDATE&sortOrder=descending

Instead of bypassing queries all together, our Info Hubs are meant to provide a landing page with links to content--blogs, discussions, guides, videos, white papers, etc.--spread across different channels and related to a specific topic. Ultimately, we want to make our content easy to find, so we curate it on these Info Hubs. For example, certain types of content we publish to EMC Online Support, such as KB articles and patches. Several teams at Isilon publish white papers for different audiences, so we try to track those white papers on Info Hubs as well. We understand that publishing to many places makes it difficult to track content, so we hope Info Hubs can help.

Speaking of a Wiki, the hubs are editable by everyone -- is this meant to encourage all of us to participate?

A key reason we wanted to do Info Hubs is because they are interactive. So yes, we encourage participation and welcome feedback through comments. We frequently edit them to keep them updated with new content, but if you find a piece of information you would like to include, we welcome your edits.

In regards to your feedback about the EMC Online Support site, this is managed by the EMC Customer Service team. They have a support blog and community: Connect with EMC Customer Service.This could be a good place to ask questions and share feedback. 

In regards to the Overview page on the Isilon Community, we use the same template as other product communities, but I think there is area of opportunity to remove redundant widgets and funnel people to forum discussions based on their level of expertise or topic of interest. We're working with our Isilon Community Manager to evaluate how to update this Overview page. This will probably be an ongoing process where we make improvements over time.

As for actual documention content: In fact and authentification and network have something in common, they need to be planned and set up with teams outside storage people. The Isilon authentification White Papers can hardly be consumed by AD/LDAP admins, just to specific on Isilon CLI examples, which doesn't buy them much. A higher level description and graphics of concepts, mechanisms and scenarios that can be shared by Isilon *and* AD/LDAP admins can foster communication and avoid misunderstandings.

This is an excellent point! In regards to authentication white papers, are you referring to "Identities, Access Tokens, and the Isilon OneFS User Mapping Service," for example? I'll submit this feedback to isicontent@emc.com. This is an email distribution list that includes Isilon content creators from other teams, such as my team (Information Development), Product Marketing, and Training. If you have additional content ideas, you can use this email. (I include this email in the footers of most ISI Knowledge blog posts, but I can share it more prominently on the Community as well.)

I'm also sharing your comments with the appropriate content developers about providing more rich scenarios in our content, and adding content to drive support packages.

Biggest issue by far for me is keeping track of OneFS software releases, patches and know issues.

Members from our team--who maintain the Current Software Releases doc, the Release Notes, and ETA KB articles--had a discussion about this on Wednesday. We identified some ways that we can make it easier to track resolved issues across release notes and ETA KB articles. We're also in the process of developing a Patch Info Hub. This information is very complex to track, so we're dedicated to make this information more clear. I also like your suggestion to add "Open/Resolve" to the ETA hub, and we've taken this into consideration.

I like the new format of unified family and MR releases. Btw, with a new minor family like 7.2.1, what are the initial "fixed issues", fixed since 7.2.0 or some specific 7.2.x?

Glad to hear you like the redesign! The 7.2.1 release notes include fixes issues from 7.2.0.2 and 7.1.1.4 (on page 17).

Frankly, I can't get around this. Where is something "otherwise noted"? What kind of Big Data analysis am I supposed to perform by hand here, in order to find out which OneFS release will be fit for my environment? I would need to cross-reference the notes with the available patches, too, of course. Suggestion: Please (please) publish a table (Excel, csv) organized by issue id and code family, listing the issue status with the family, e.g.: Issue 99999  | OneFS 7.1.1 | fixed in patch 5555 for 7.1.1.3, fixed in 7.1.1.4 Issue 99999  | OneFS 7.2.0 | unresolved

We definitely want to make it easier to find an answer to the question: will OneFS fit my environment. I like your suggestion about single sourcing all this information into a sortable table, and we're exploring the how we can build this type of table in a way that it can be quickly updated with relevant information.

I really appreciate hearing your thoughtful suggestions. I hope I was able to address most of your comments. Happy to continue the conversation as well. Thanks again for your notes and feedback!

Hi Everyone

1.- is there some section about updating courses?

If I took the course about Implementation Isilon with onefs 7.0

Where could I find information about oneFs 7.2 , no release documentation, I would want information with more details about news of ones 7.2, for example, if deduplication is the new feature  in the isilon, I would want to find a document that it explains how to work deduplication

does it exist?

Thanks

Hi Kirsten

thank you, I really appreciate your detailed response and look forward to upcoming additions and improvements.

And yes that AIMA White Paper is at the core of what I had in mind.

Cheers

-- Peter

Hi Francisco,

If you are looking for documentation about OneFS 7.2 features and how to use them, you can find content links on this Info Hub: OneFS 7.2 Documentation - Isilon Info Hub Specifically, the OneFS 7.2.0 release notes will summarize new features in this release, and you can read more about those features in the OneFS 7.2 Web Administration Guide.

Are you also looking for information about OneFS administration courses?

Thanks,

Kirsten

Thanks, Peter! I'll pass the feedback about this white paper to our isicontent@emc.com email address about addressing AD/LDAP audiences.

This Ask the Expert event has officially ended, but don't let that retract you from asking more questions. At this point our SME are still welcomed to answer and continue the discussion though not required. Here is where we ask our community members to chime in and assist other users if they're able to provide information.

Many thanks to our SMEs who selflessly made themselves available to answer questions. We also appreciate our users for taking part of the discussion and ask so many interesting questions.

ATE events are made for your benefit as members of ECN. If you’re interested in pitching a topic or Subject Matter Experts we would be interested in hearing it. To learn more on what it takes to start an event please visit our Ask the Expert Program Space on ECN.