IT-events | 中国 IT 活动日历 | Crawler library

In my project I'm using Jhipster Spring Boot and I would like to start 2 instances of one microservise at the same time, but on different instances of a database (MongoDB).

In this microservice I have classes, services, rests that are used for collections A, B C,.. for which now I would like to have also history collections A_history, B_history, C_history (that are structured exactly the same like A, B, C) stored in separated instance of a database. It makes no sense to me to create "really separated" microservice since I would have to copy all logic from the first one and end up with doubled code that is very hard to maintain. So, the idea is to have 2 instances of the same microservice, one for A, B, C collections stored in "MicroserviceDB" and second for A_history, B_history, C_history collections stored in "HistoryDB".

I've tried with creating 2 profiles, but when I start from a command line History microservice, it is started ok, but if I also try to start "original" microservice at the same time, it is started but immediately history service becomes "original" microservice. Like they cannot work at the same time.

Is this concept even possible in microservice architecture? Does anyone have an idea how to make this to work, or have some other solution for my problem?

Thanks.

application.yml

I want to use wolkenkit's eventstore and was trying to set up a quick example. But I'm not able to simply output an event stream.

Simplified example:

I'm trying to convert a password to a sha1 sum before submitting. I'm using crypto.subtle.digest to do the conversion. Since it returns a promise, I await on it. The problem is, that the password is submitted unconverted.

I have some spring-boot micro-services deployed on Cloud Foundry and I have to implement propagation and storing (to a repository) the business audit events emitted by them.

I can do it in several ways, e.g.:

1. Publish audit events to a Source (Spring Cloud Stream / RabbitMQ) and consume it by a Sink service writing the events to the repository.
2. Publish audit events as a custom application log and consume it by a log-consuming service filtering and writing the events to the repository.
3. Publish audit events using internal CF's mechanism as a new custom log type or a custom audit event - I think this option isn't a good idea but I can be wrong...

Is there any recommended approach/pattern to realize this issue on Cloud Foundry platform?

EDIT

All the approaches meet (in my opinion) the 12-factor rules, but each has its advantages and disadvantages:

• (1) Spring Cloud Streams
  • + ensures delivery (events will not be lost)
  • + allows to use routing (RabbitMQ)
  • - requires connection to a message broker (not as easy as a logger)
• (2) log-consuming service
  • + is easy
  • - logs can be lost
  • - audit biznes info is too widely propagated - GDPR
• (3) new CF's log type
  • - probably forces changes in the CF

I can see the Repository level Audit log in the Repository Settings, and I can see the Project level Audit log in the Project Settings.

What about the "System" level audit log which contains Server, User Management, Permission events, etc., per https://confluence.atlassian.com/bitbucketserver/audit-events-in-bitbucket-server-776640423.html?

I am on version 5.1, and I have "System Admin" permission, but I do not have access to the logs on the server (i.e. /log/audit). Also, I exported the support zip via "Support Tools" with all application logs enabled, and it does not appear to contain the audit logs.

Is there a possibility to capture an exit or quit in ST3? I saw this older question here and I tried both the on_windows_command and the on_text_command but neither seems to trigger the quit/exit!?

It there's none, it would also be fine if I could handle it on a restart of sublime, but the on_load doesn't seem to be called again on remembered views!?