Why isn't Asset Management more developed?

[Monty]

When I first saw the Asset Management feature, my mind was blown. Like, I'd have given a kidney to have a tool like this back in some of my previous jobs.  But then I started actually using it and I'm finding it pretty limiting.  Like, this could be doing a lot of heavy lifting in JSM, but it just doesn't seem to be very well integrated.

Some examples:

• Unable to set default values or hide the field in in a request type.
  • [JSDCLOUD-11227] Default value for object attributes - Create and track feature requests for Atlassian products.
  • [JSDCLOUD-10317] Ability to hide Assets object custom field in customer portal - Create and track feature requests for Atlassian products.
  • This seems like essential functionality, and the "workaround" really doesn't solve the problem.
  • This one thing could solve a whole lot of problems for folks.
• Unable to use a user's email address in place of the account ID for user attributes.
  • [JSDCLOUD-10487] Allow to use user email address when import user attribute in Insight - Create and track feature requests for Atlassian products.
  • I'd argue, that this should be a default option everywhere, not just for import.  I'd want to use this in JQL/AQL as well.
• Atlassian's own Entra ID integration plugin only supports a manual import with no apparent way to schedule or trigger it.
  • How to automate the import of EntraID users with J... (atlassian.com)
  • Why wouldn't they add a scheduling feature to this?!  In what realm would an organization choose to import from Microsoft just one time and then never again? Of course you'd need to keep the objects synchronized regularly, so why not give us the ability to schedule the import?
• The documentation is horrendous.
  • Here's an interesting trek. 
    • Step 1: Automate your objects in Assets | Jira Service Management Cloud | Atlassian Support 
      • This page has no useful information and points you to Learn more about automation for Jira.
    • Step 2: Learn more about automation for Jira. Well, I guess we go to Get Started?
    • Step 3: View the guide
    • Step 4: Give up I guess?
  • Install an asset management app | Jira Service Management Cloud | Atlassian Support
    • Why do I need an Asset Management App? Is there a way to filter for an Asset Manager App in the marketplace? What even is an Asset Manager App? Is it required? 
• What is the Assets objects field? | Jira Service Management Cloud | Atlassian Support
  • "The Assets objects field functions differently in Cloud than it does in the Server/DC version of Assets. In Assets for Cloud, users cannot use an Assets object field to sort search results in a JQL query."
    • WHY?!?!?! You are pushing people to the Cloud where some important features are just not implemented!
    • You can't use non-asset fields in a `aqlFunction(aql)` it only works with asset object custom fields.  That means if you can't use things like `currentUser()` in an aqlFunction.
      • You can do things like: 

        • Find issues where a website is reported and where the customer account manager is Mathias Edblom:
          customField in aqlFunction("Customer.\"Account Manager\" = \"Mathias Edblom\"")

      • But you can't do things like: customField in aqlFunction("Customer.\"Account Manager\" =currentUser()")

So, back to my question, why isn't Asset Management more developed? I really want it to be good, and it's so close, but I keep running into roadblocks over and over again. It's making it challenging to advocate for this product at my organization.

[Nic Brough -Adaptavist-]

Welcome to the Atlassian Community!

Assets is something Atlassian acquired, relatively recently, not wrote, and they are trying to integrate into JSM as best they can.  The original product was not closely aligned with the way Jira does things, so it is slow going.  (But the original product worked as well as it could given everything had to be done outside Jira)

I know that is a very apologistic thing to say, but the short answer is "time".  Atlassian's Assets team simply has not had enough time to develop everything that we want.

[John Funk]

Well said

[Monty]

Ok, but they keep choosing to acquire new things that take time to integrate when they haven't finished integrating the last 10 things they acquired.

If they are going to acquire a product, they should allocate the resources to integrate it fully. Instead, they have this weird mashup of products that kind of work with each other, but everything has a unique licenses, and nothing is ever finished.

They are gung ho about Compass and Atlas, but still haven't finished integrating Assets.  I get that it's hard, and it takes resources, but they are making the decision to spread those resources thin.

[Monty]

As I said in my original post, this product has SO much potential. I advocated STRONGLY for buying the additional licensing to gain access to it. I'm now on the hook for showing its value and I'm struggling.

I can make some things work, but the places where integration isn't fully baked is alarmingly bad.

Here's an example. I can create an object with an attribute that connects the jira user account with the object. (It's even in their People template).

Great, so I have an object, and a user account is attached to the object, but then I can't use that asset in any place where a user would be useful.  I can't assign tickets to that User's asset object. It doesn't automatically build relationships with that the jira account without creating a custom asset field (which I can't set a default for or hide).

So now, I have to create a custom asset field, then use automation (which is limited) to try to copy the assignee or reporter to the asset field when it should already be linked.  From the issue view, now we have multiple fields that an agent has to see (and not touch) so that everything works.

Here's a real world example that I'm trying to solve:

An agent needs to escalate an issue to their manager.

We have an asset object called Person that has attributes for jira user, and manager (another Person object).

So, I have all the relationships I need right? I have the assignee (jira user), that jira user is in an attribute of a person object, that object has a relationship with another person object for their manager, and the manager person object has a relationship with its own jira user. 

But I can't use any of those relationships because aqlFunction() can't use non-asset functions to lookup values.

I should be able to create a Queue with a JQL/AQL that says: If one of my employees raises an issue, put it in this queue.

[Nic Brough -Adaptavist-]

No sane organisation would stop everything else they are doing in order to consolidate and integrate one thing that they have acquired. 

An organisation like Atlassian has dozens of products and services, and they have aspirations to do grow and do more.   You are asking 10,000 people to stop doing anything until a new thing has been developed to the point that a customer is demanding, which is frankly silly, as most of the organisation does not have the skills to do anything to help.

Would you expect a supermarket to tell you "we're only going to sell you fruit for the next month, while we roll out our new dragonfruit product"?

[Monty]

I don't think that the entire company should (or could) be engaged on a single product feature, nor should they work solely on a single thing at a time. It's a straw man's fallacy to misrepresent what I said by exaggerating it to that extreme.

If your argument is that the main reason that the product is in the current state is that they need more time, then I'd argue that their Roadmap does not indicate that they have any plans for fixing any of the things that I've talked about. The only items in the future or coming soon status related to assets is a Reporting feature, support for more objects, surfacing assets in Confluence, and incorporating yet another acquisition called Air Track to analyze assets. Ok, so, we've got an assets team that is working on Assets, but are they working on the things that people are asking for? I'd argue that their focus isn't on further integration with Jira and no amount of time will change that. It seems like they got it to a "good enough" state and moved on.

What I'm talking about is integrating it with their existing product so that assets can be used more easily. Assets could be used to drive a lot of the data objects in their product. I'd love to be able to use an Asset instead of "Teams" in a lot of places for instance. If I have all my relationships mapped, why do I need a separate thing for Teams?  Because they are part of two different products that aren't integrated well enough to talk the same language.

[Nic Brough -Adaptavist-]

It is however what you did suggest.

But that does not change the point - the Assets team have a lot of things they would like to do, a lot of things that customers want them to improve, and simply not enough time to do it all now.

[Patrick]

The worst thing is that after implementing it is that they are not even part of any backups. There is not even a bin where you can restore from. 

Thanks, Atlassian once again, forcing me to either manually export it as tables?? or having to buy some third party plugin. Its a joke.

Sorry, just wanted to add some of my frustrations here as well. 

[Vish Reddy _Revyz_]

Hi @Jingis, Patrick, @Monty ,

I represent Revyz, While I can't solve all issues in using Assets, I can provide with our Assets Data Manager app for free for you for a period of a year if you are open to helping us improve it further, atleast it will help automate some tasks for you, I am by no means saying it will solve all the issues you have highlighted.

As of today we automate the backup and granular restore of Asset config and objects. We have a roadmap which addresses additional use cases and I would love to work with you to see what we can do further.

Please drop me an message (vish@revyz.io) if you are interested.

Thank you

Vish 

[Monty]

Wait, what do you mean it's not backed up?!

So, we put all our asset information into this system and if something happens there's no way to recover or restore anything? 

Oh gosh. What if someone comes in and mucks it all up? That's one mistake away from a major problem.

[Vish Reddy _Revyz_]

Hi @Monty 

Part of the Cloud Security Shared Responsibilities is that a customer is responsible for their data that lives in the SaaS application.

This is true with all SaaS applications, be it Microsoft, SalesForce, Atlassian etc..

Nothing unusual about Atlassian in this context i.e. Atlassian is no different from other SaaS vendors.

Here are links to some of the leading companies Cloud Security Shared Responsibilities docs:

• Microsoft / Azure
• Salesforce 
• AWS 
• Atlassian

From an Atlassian perspective you should review these two docs:

1. Atlassian Cloud Security Shared Responsibility Model

2. Atlassian Security Practices

As you would see in the Atlassian Security practices doc -

"We do not use these backups to revert customer-initiated destructive changes, such as fields overwritten using scripts, or deleted issues, projects, or sites. To avoid data loss, we recommend making regular backups."

If someone on the customer side mucks up something it is the responsibility of the customer to be able to recover back their data and hence the need for a customer to backup their data.

In the most recent example from yesterday, Atlassian did an upgrade / change to the system which messed up Automation rules for a large no. of customers, in that scenario Atlassian was responsible for getting things back to normal (as they did or are in the process), on the other hand if someone on the customer side deleted something inadvertently (which of course happens) then it is the responsibility of the customer to have backup their Automation Rules (in this case) to recover back!

I am just the messenger here :-) 

[Patrick]

Sure, I do the backups gladly, which can be automated easy in AWS or Microsoft. But the problem I'm trying to highlight is that they don't even offer a proper backup system?! At least, I mean, at least have a bin if someone deletes something by accident. 

Don't even get me started with Confluence Backups, they are the worst thing I've ever came across. You have a 10GB Instance, good luck bro, there is a 200MB upload limit if you want to restore something. Just Host a Datacenter, import the backup there and then use a Cloud Migration tool.
Ahhm what??! 

Thanks for the offer for the plugin, will take a look. It's the only option anyway, to use a 3rd party plugin.

[Monty]

<deleted>