COMDEX namespace environment variables not being applied about cosmos-omnibus HOT 9 CLOSED

@ourlink the environment variables don't change the actual files, they just take precedence over the values in the config files.

The value for API_ENABLE is true not 1 as per the config files. I suspect that's all you need to change to fix this issue.

from cosmos-omnibus.

Okay, I understand now. The files do not change it is the ENVIRONMENT VARIABLES that are passed to the docker instance at startup.

However, I tried all the variations of the following and none worked.
COMDEX_API_ENABLE=True
COMDEX_API_ENABLE=true
COMDEX_API_ENABLE=TRUE

During my testing, I tried different providers to ensure it was not the provider having network issues.

None of these settings enabled the API port 1317 although the port was mapped and presented with a redirect.

Any ideas?

I have pasted my test SDL below

---
version: "2.0"
services:
  node:
    image: ghcr.io/akash-network/cosmos-omnibus:v0.4.4-comdex-v13.4.0
    env:
      - MONIKER=ValidatorNode_Sentry1
      - CHAIN_JSON=https://raw.githubusercontent.com/cosmos/chain-registry/master/comdex/chain.json
      - STATESYNC_RPC_SERVERS=https://comdex-rpc.polkachu.com:443,https://rpc.comdex.one:443,https://rpc-comdex.goldenratiostaking.net:443
      - STATESYNC_SNAPSHOT_INTERVAL=1000
      - MINIMUM_GAS_PRICES=0.025ucmdx
      - P2P_POLKACHU=1
      - STATESYNC_POLKACHU=1
      - COMDEX_API_ENABLE=true
      - COMDEX_P2P_PRIVATE_PEER_IDS=9e1ee53832d330adcaf0f9830e2b4ee1be6b6123
      - COMDEX_P2P_UNCONDITIONAL_PEER_IDS=9e1ee53832d330adcaf0f9830e2b4ee1be6b6123
      - PRUNING=custom
      - PRUNING_KEEP_RECENT=100
      - PRUNING_KEEP_EVERY=0
      - PRUNING_INTERVAL=10
    expose:
      - port: 26657
        to:
          - global: true
      - port: 26656
        to:
          - global: true
      - port: 1317
        to:
          - global: true
profiles:
  compute:
    node:
      resources:
        cpu:
          units: 2
        memory:
          size: 10Gi
        storage:
          size: 125Gi
  placement:
    dcloud:
      pricing:
        node:
          denom: uakt
          amount: 1000
#      signedBy:
#        anyOf:
#          - akash1365yvmc4s7awdyj3n2sav7xfx76adc6dnmlx63
      attributes:
        host: akash
deployment:
  node:
    dcloud:
      profile: node
      count: 1

from cosmos-omnibus.

I use this config (COMDEX_API_ENABLE=true) on one of my public nodes so I'm sure it works, I suspect it's just a port mapping issue. Have you deployed the test SDL on Akash currently, and do you have the mapped port I can test? Your SDL looks fine.

from cosmos-omnibus.

Yes, I have it (above SDL) deployed and the current mapping for port 1317 is http://provider.validatornode.com:31787/

BTW - I have tried deploying this SDL to other providers and I get the same results.

from cosmos-omnibus.

Could you try setting the following additional config:

COMDEX_API_ADDRESS=tcp://0.0.0.0:1317

The default may have changed to use a different port or be more restrictive.

from cosmos-omnibus.

Good Catch!
That change appears to have resolved the issue.

from cosmos-omnibus.

Letting the chain re-sync and I'll know for sure in 15 minutes.

from cosmos-omnibus.

Issue has been resolved!

from cosmos-omnibus.

Great news, thanks for confirming @ourlink

from cosmos-omnibus.