api-layer icon indicating copy to clipboard operation
api-layer copied to clipboard

Incorrect components displayed on Eureka page when disabled 1 lpar on sysplex.

Open ZainSidd opened this issue 3 years ago • 1 comments

Describe the bug Working on HA Sysplex component availability when 1 instance is stopped. Have 3 instances up. 2 are from the same machine while the other is on a different machine, i.e lpar1, lpar2a, lpar2b.

Steps to Reproduce

  1. Start three instances of zowe on sysplex, 1 on one machine, 2 on one machine.
  2. Once all instances have started and all components report correctly on the eureka page turn off the main instance, the one that is 1 on one machine.
  3. Wait a little bit for the instance to turn off, then go to the Eureka page for lpar2a.
  4. There it will be seen the wrong number of components available being reported.

Expected behavior What is expected is to see on the eureka page for lpar2a is that lpar1's components disappear from the dynamic components and you are left with 2 components in each dynamic component. What occurs is lpar2a shows lpar2b's components and vice versa for lpar2b on the eureka web page. https://github.com/zowe/zowe-install-packaging/wiki/Test-Zowe-in-Sysplex. In section D of that webpage, it says what the steps are and the expected result, the difference in this test is we are using 3 instances instead of 2 in section D.

Logs If applicable, add server logs collected at the time of your problem.

Details

  • Version and build number: v2.0.0
  • Test environment: ZHAPLEX, Tivlp13, Tivlp14

API Catalog Web UI (in case of API Catalog issue):

  • OS: [e.g. macOS, Windows]
  • Browser [e.g. Chrome, Safari]
  • Version [e.g. 71.0.3578.98]

REST API client (in case of REST API issue):

  • Technology: [e.g. Spring Boot, Node.js]
  • OS: [e.g. Windows 10]

Additional context Add any other context about the problem here.

https://github.com/zowe/api-layer/wiki/Issue-management

ZainSidd avatar Jun 22 '22 19:06 ZainSidd

HA Logs.zip Some logs from the when i try to do the above. The logs are after 1 LPAR is taken down.

ZainSidd avatar Jun 29 '22 18:06 ZainSidd

@ZainSidd from the logs that you have provided it looks like you have a connection issue between instances, can you provide logs from instances that are correctly registered?

achmelo avatar Jan 09 '23 16:01 achmelo

Since there is no feedback we are closing the issue. If the problem does persist, feel free to reopen.

anton-brezina avatar Feb 01 '23 13:02 anton-brezina