One advantage of the old Service Manager GUI ("Dialog Client") was that it had a back-channel for communicating with processes it started. This allowed it to collect additional information such as the AE server process name, AE server role, and number of connections.
When running the Automation Engine in Kubernetes (AAKE), there is no Service Manager, and Kubernetes essentially fills this role. However, the additional information previously available via the Service Manager GUI is not easily accessible to the Kubernetes cluster or the people operating it.
I suggest that Broadcom change this! I propose that AAKE be enhanced to use Kubernetes labels to store the additional information. This includes information previously collected by the Service Manager, but also some new suggestions.
- AE server name (AE_EXP1)
- Process name (WP001)
- Process type (WP, CP, JWP, JCP)
- Process role (PWP, DWP, OWP, RWP, WP, CP, JCP-WS, JCP-REST, JWP, JWP-AUT, JWP-IDX, JWP-UTL, etc.)
- Network area
- Number of connections
This could be done in a number of ways, including:
- Introduce an additional process that collects this info and patches the pods periodically.
- Update the AE server programs so that each process is able to patch its own pod.
If you want to be really fancy, you could also use annotations to store additional non-identifying metadata such as log file path or time of last role change.
The above enhancement would make AAKE troubleshooting much easier, and would enable using K9s as a modern version of the old Service Manager GUI.
Context: ae_exp_1 [RW] <0> all <a> Attach <ctrl-k> Kill ____ __ ________
Cluster: 086ad919-520b-4a69... <1> uc4 <ctrl-d> Delete <l> Logs | |/ / __ \______
User: 7d26dc0d-60f0-4333... <2> default <d> Describe <p> Logs Previous | /\____ / ___/
K9s Rev: v0.50.6 ⚡️v0.50.12 <e> Edit <shift-f> Port-Forward | \ \ / /\___ \
K8s Rev: v1.29.6+vmware.1 <?> Help <z> Sanitize |____|\__ \/____//____ /
CPU: 2% <shift-j> Jump Owner <s> Shell \/ \/
MEM: 19%
┌──────────────────────────────────────────────────────── pods(uc4)[71] ────────────────────────────────────────────────────────┐
│ NAME PROC TYPE ROLE CONN PF READY STATUS RESTARTS CPU MEM %CPU/R %CPU/L %MEM/R %MEM/L IP │
│ jcp-rest-0-77bc8664c9-94m99 CP001 JCP REST 1 ● 1/1 OK 0 2 177 2 0 24 9 10.42.1.11 │
│ jcp-rest-0-77bc8664c9-5vfwk CP002 JCP REST 1 ● 1/1 OK 0 4 176 2 0 23 11 10.42.1.12 │
│ jcp-ws-0-54668588cf-d6fwf CP003 JCP WS 0 ● 1/1 OK 0 4 209 2 0 25 9 10.42.1.13 │
│ jcp-ws-0-54668588cf-9wbnx CP004 JCP WS 4 ● 1/1 OK 0 1 175 1 0 24 10 10.42.1.14 │
│ jcp-ws-0-77bc8664c9-sgk5z CP005 JCP WS 4 ● 1/1 OK 0 1 176 1 0 25 9 10.42.1.15 │
│ jcp-ws-0-54668588cf-4trwd CP006 JCP WS 10 ● 1/1 OK 0 2 177 2 0 24 9 10.42.1.17 │
│ jcp-ws-0-54668588cf-4trwd CP006 JCP WS 8 ● 1/1 OK 0 2 177 2 0 24 9 10.42.1.18 │
│ wp-0-b858d895d-xgtfn WP001 WP PWP* 0 ● 1/1 OK 0 2 178 1 0 24 9 10.42.1.19 │
│ jwp-0-6c78445657-skrmn WP002 JWP AUT 0 ● 1/1 OK 0 2 409 2 0 24 9 10.42.1.20 │
│ jwp-0-6c78445657-r4btv WP003 JWP IDX 0 ● 1/1 OK 0 2 220 1 0 23 9 10.42.1.21 │
│ ... ... ... ... ... ... ... ... ... ... ... ... ... ... ... ... │
└───────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────┘