I'm working with the Okta API and trying to understand a couple of things:
How do you get a count of members. For example, how do I get a count of groups defined, or the count of users assigned to a group. The admin tool seems to do this easily enough, but I need to do this in another tool I'm working on.
I can iterate through the pages until I get to the last page, incrementing a counter as I go along, but this is very slow, especially when trying to get the number of groups defined for my subdomain, or other similarly high-numbered value.
How do filters and filterbuilders work? The basics are pretty simple, but the filter syntax doesn't seem to be defined anywhere.
Thanks!
AFAIK the only way to get a count is by getting all users/groups and subsequently the count. There is no API call to get a summary. Basic, but works.
Check the Okta API docs at http://developer.okta.com/docs/api/resources/users.html#user-model for the user model. You will get a collection of those JSON objects when you list the users - http://developer.okta.com/docs/api/resources/users.html#list-users
Call this last API and count the instances that you get back in the JSON collection.
Related
Given a unique_email_id, how can I retrieve all list members matching that ID?
I am attempting to use the search-members endpoint, ala what I do for searching email-addresses:
https://us9.api.mailchimp.com/3.0/search-members?query=chris#chris.com
but instead (given a unique_email_id of 4dce5
https://us9.api.mailchimp.com/3.0/search-members?query=4dce5
and I'm getting no results.
Is there another endpoint I'm missing?
However using the query string ?unique_email_id=X attached to the end
of GET lists/list_id/members will pull up anyone with that specific
EUID.
Indeed this works very well. Example :
GET /3.0/lists/dfc3cf4b38/members?unique_email_id=f4b75a9d78&fields=members.email_address
will return :
{"members":[{"email_address":"my_email#hotmail.com"}]}
I've had a chat with the mailchimp support about this issue. This was there response (part of it because not everything discussed in that chat is relevant to this question):
(09:00:04 AM) Mailchimp support: Thanks for hanging in there, Me. It looks like
the email_unique_id can't be searched via API endpoint like it can be
within the app. Another option to get to your goal is GET to
lists/list_id/members with the query string
?fields=members.email_address,members.unique_email_id which should
pull up all their subscribers' emails and unique ids. At that point,
you can filter through those results for the unique id
(09:00:52 AM) Me: Same goes for the user ID I guess?
(09:02:05 AM) Me: Why can't it be searched? When I list ids as well I might miss information. I don't need the unique_email_id in my response, I need to search for it to retrieve user information without getting duplicates in my result
(09:07:11 AM) Mailchimp support: Ok I understand, one
moment while we test this out for you. Thanks again for your patience!
(09:16:22 AM) Mailchimp support: Hey Me, thanks for waiting. I tested the
endpoint lists/list_id/members?unique_email_id=e8da8fa60a and I was
able to return just the one subscriber with that field. Search-Members
only functions the way it does in app where you can only search
specific strings like names, email addresses, but not ids as those
aren't visible in app. However using the query string
?unique_email_id=X attached to the end of GET lists/list_id/members
will pull up anyone with that specific EUID.
I have an Asp.net Core application I want to be able to allow multiple/ different Tenant(Client)to access the same application but using different url's. I have common database for all tenant(client).
So It is the main part I want to host my application in a domain say... www.myapplication.com then allow different Tenant(client) to access the same application using
1.www.TenantOne.myapplication.com
2.www.TenanatTwo.myapplication.com.
3.www.{TENANCY_NAME}.myapplication.com
I can't find any info on how to do this and I'm stuck.
How to do it? Please provide the code. Thanks.
As Saravanan suggested these types of questions don't belong here on SO. To get you started, I suggest you start looking if there are any frameworks such as SaaSKit available to add a multi tenancy layer to the pipeline.
The essential part is to know where each request comes from. Using subdomains is a good way to achieve that and middleware is a good place to 'identify' your tenant. You could have a database to persist the tenants but the implementation is entirely up to you. I also wrote a little article on the subject. Although it isn't ASP.NET Core, the principles still apply.
The approach I believe you are looking for is similar to the article at the url below.
https://dotnetthoughts.net/building-multi-tenant-web-apps-with-aspnet-core/
In it, the author splits the requesting URL into an array of strings delimited by the dot in the address. The variable 'subdomain' is then set to the first element of that array. In your question, it looks like you may want to use the second element in the array, but you get the idea.
var fullAddress = actionExecutingContext.HttpContext?.Request?
.Headers?["Host"].ToString()?.Split('.');
var subdomain = fullAddress[0];
//do something, get something, return something
How you use this data is up to you. The author of the article created a filter attribute, but there are many possibilities such as passing the tenant name as a parameter to a service function.
Sorry,you have to get something to start with and then come back for the people to help you with.
I would say that this is all of a domain based wild card mapping and change in your authentication logic to get the tenant id from the URL. Once you identified the tenant, you just login and then take it forward. Like you might be having a database with the tenant details like
tenant1 | tenant1.company.com | guid-ofthe-tenant | etc...
Once you get the URL, you lookup in the above table and get the tenant code and then you choose the login mode and then proceed.
In case you have tried something yet, we would be happy to point you if it does not work yet.
I can view OKTA users in a group (/groups/{groupid}/users), and I can grab all groups belonging to a user (/users/{userid}/groups).
But both of those can return a lot of data and the latency is too high because of it. Can I specify which fields to return in the second call above?
How can I do a simple check if a user is a member of a group?
Okta Expression Language contains group functions such as isMemberOfGroup, but there is no examples or explanation of how to use that as part of an API call. I tried using it with the filter querystring, but no go.
Unfortunately you cannot as the Okta API does not currently offer an isMemberOfGroup method. You'll have to use either of the 2 APIs you mentioned in your question, whichever is the most efficient in your specific case.
I am using a filter to apply object level permissions to a collection. Resources in a second collection have a many-to-many relationship with the first. On the browsable API, when creating resources in the second collection, the user is presented with a list of resources from the first to link it to. However, this list is not filtered, so the user can see values that they should not be able to see.
I've poked around the documentation and source a bit and I cannot see a way to add filtering to the queryset that generates the choices without overloading or modifying a bunch of code to pass the request data down (probably removing some of the collection specific data on the way) and then apply the filters.
Is there a better way to achieve this?
Currently there's nothing to support this out of the box. Pull requests are always welcome. If it's something you want to work on you may want to either open a ticket on GitHub or hit up the mailing list to discuss it first.
I am able to get a list of Campaigns and filter on those to get only the campaigns I want.
How do I get a list of Contacts for 1 specific campaign?
I have the web service classes in my project but can't see any way of relating to the Child objects of the Campaign object.
Documentation is awful, none of the examples had working code and they needed amending in order to work.
Answer is don't use v2 of the API - use v1 instead which means no strongly typed objects etc but it's the only way to do it