Apps don't restart or restage in isolation segment
Article ID: 298403
Updated On:
Products
VMware Tanzu Application Service for VMs
Issue/Introduction
Restarting and restaging apps in isolation segment fail with error "All instances crashed" and "Start unsuccessful" respectively but all VM's are up and running.
cf restage <app-name> -v shows that after request "POST /v3/builds", all the following requests are "GET /v3/builds/xxxxxxx-yyyyyyy-zzzzzzzzzzzzz" with "droplet": null" in the response body, and also "GET /api/v1/read/aaaaa-bbbb-cccc-dddd-eeeeeeeeeeeeee" which collects the logs.
The above means that the staging is blocked at the building phase, which is done by Diego Cells and requested by Cloud Controller . This happens if Cloud Controller VM's in TAS deployment can't connect to Diego Cells in Isolation Segment deployment.
You can "bosh ssh" into Cloud Controller VM's and check connectivity to Diego Cells in Isolation Segment with ping or "nz -zv" on port 1801 (which is rep process) to confirm the above.
cf restage <app-name> -v shows that after request "POST /v3/builds", all the following requests are "GET /v3/builds/xxxxxxx-yyyyyyy-zzzzzzzzzzzzz" with "droplet": null" in the response body, and also "GET /api/v1/read/aaaaa-bbbb-cccc-dddd-eeeeeeeeeeeeee" which collects the logs.
The above means that the staging is blocked at the building phase, which is done by Diego Cells and requested by Cloud Controller . This happens if Cloud Controller VM's in TAS deployment can't connect to Diego Cells in Isolation Segment deployment.
You can "bosh ssh" into Cloud Controller VM's and check connectivity to Diego Cells in Isolation Segment with ping or "nz -zv" on port 1801 (which is rep process) to confirm the above.
Environment
Product Version: 2.13
Resolution
Check new firewall or nat rules added to network that can cause traffic to be dropped when going to the Isolation Segment subnet. Deleting this rules should solve the issue.
Feedback
Yes
No
Powered by
