WSO2 as a Orchestration Service or Apache Nifi as Orchestration Service - wso2

I am new to WSO2 and Apache Nifi. After having gone through official WSO2 and Apache Nifi documentation and some other sites, I am still having doubts regarding my use case and using WSO2 or Apache Nifi for that.
My use case is as follows
I have multiple REST APIs, each of which is a processing step in a multi-step processing. Some of the processing steps can take time in the magnitude of seconds (2 to 4s). In order to scale them, I want to run multiple instances of each processing steps (REST APIs) on same port but different VM.
In order to orchestrate call between these processing steps, What is more appropriate to use Apache Nifi or WSO2 ESB Service Orchestration.

Related

multiple tenants accessing Confluent kafka on GCP

For multiple external tenants accessing Kafka, is there any issue with providing the same endpoint (ie. set of brokers, ports) to the multiple producers ?
what are the best practices wrt multiple tenants producing data on (Confluent) Kafka topics, Kafka being installed on GCP.
tia!
Here are Google's Best practices for enterprise multi-tenancy
Confluent Cloud is a fully managed Apache Kafka as a Service in the public cloud. Looks like you can register here to download the white paper that contain a guide to Kafka best practices.
As well you may find this presentation interesting.
Apache Kafka is available throw Google Market Place.
Confluent is a Google partner and you can reach out to them here by clicking on the Contact partner button and filling the form.

WSO2 API Manager Analytics Accuracy

We've deployed WSO2 Api Manager in Active-Active scheme with single Api Manager Analytics server. Also we're using Graylog+Elastic+Grafana to collect logs from back-end services behind API.
The case is that amount of back-end services calls are hugely greater (millions vs thousands) than a corresponding API hits count. Why is that? What are we missing?

WSO2 Identity Server clustering: does each node in a cluster need to share a common SQL database?

We are evaluating WSO2 Identity Server, and are trying to understand how WSO2 Identity Server scales across multiple nodes.
The document https://docs.wso2.com/display/CLUSTER44x/Clustering+Identity+Server+5.1.0+and+5.2.0 explains steps to follow to enable clustering, however the document links to database setup documentation that is common to the whole WSO2 suite, and not specifically for Identity Server.
I understand from the documentation that all WSO2 Identity Server nodes are required to share a set of clustered LDAP servers (when an LDAP server is in use).
Is each node in a WSO2 Identity Server cluster required to share a set of clustered SQL servers in addition to the above?
Alternatively, is it fine to have each node use a local database in the expectation that WSO2 Identity Server will communicate with other members of the cluster in order to keep all databases in sync?
When clustering wso2 products, you can use a local db for WSO2CarbonDB (This will be used to store instance-specific data) But you have to share registrydb, userdb among all the nodes.
AFAIK wso2 will not do the db syncing if you use separate db for each node.

Micro service management

We are developing a merchant application in that we have various modules like Schedule, Booking, Invoice e.t.c, each of this module are runs in different server, those are exposed through as RESTful granular services. UI layer will communicate with these granular service accordingly. To identify the request and redirect to specific micro service runs in service layer of various sever we have created a service gateway. Some of the service required data manipulation on the go which is presently accomplished through Mule ESB and some routing activities are also managed through it.
Actual purpose of the Service gateway is to match the request with service dictionary available and redirect to the respective micro service, at present its been developed in j2ee framework and runs in wildfly server. So to achieve the same process in light weight manner we come across a micro service manager like"getKong" and Customising "nginx" server to manage microservices, Mule ESB.
Along with Service Bus management is it advisable to use the Mule ESB as MicroService maanager as like getKong or any other valuable suggestion ?
In my personal opinion, you have three options:
If you don't need to perform authentication/authorization or/and
Throttling and your routing can be quite complex/complicated than is
completely fine to do it in Mule ESB.
If you do just URL rewrite nginx is probally the best choice for
minimum overhead and maximum performances.
If you really need an API manager with all the rich features than is
fine getKong or, if you want to stay in the MuleSoft
world and your are willing to pay, you can have a look at API
Gateway.
Hope this helps

how to deploy wso2 products in aws beanstalk?

I would like to deploy wso2 products in aws using beanstalk.
Theoretically it should be possible by following the instructions for deploying in webapp mode.
http://wso2.org/project/carbon/3.2.2/docs/admin_guide.html#webappmode
Has anyone managed to achieve this?
What steps were required?
Following the document [1] you shared should work.
However, unless you really have a requirement, it is not that encouraged to run WSO2 Products in web app mode. This is because the Web app container will limit the WSO2 products, where WSO2 Products can function independently in stand-alone mode. Running WSO2 stand-alone over the EC2 might be more appropriate.
[1] http://wso2.org/project/carbon/3.2.2/docs/admin_guide.html#webappmode