search

dashpay / platform

L2 solution for seriously fast decentralized applications for the Dash network
https://dashplatform.readme.io/docs/introduction-what-is-dash-platform
MIT License
72 stars 39 forks source link

doc: which services start in which order #1745

Closed coolaj86 closed 1 month ago

coolaj86 commented 8 months ago

If you don't start the services in the right order, sometimes they just crash and don't restart (looking at you tenderdash).

Although that behavior is probably a bug... but right now I just want to document the relationships:

dashd        19999 (P2P), 19998 (RPC), 18009 (ZMQ)
├── dapi-core-streams 2510 (gRPC), <needs ZMQ>
├── drive
│   └── tenderdash 26656 (P2P, alt port for testnet?), 26657 (Tender RPC), 26658 (ABCI), 26660 (Prometheus), 36656 (seed peers?)
│       └── dapi-api 2501 (JSON RPC), 2500 (gRPC)
└── caddy (10000 testnet, 443 mainnet)
    └── envoy (3080 - arbitrary for comms with caddy)

Ports required for ProRegTx testnet:

Dashcore P2P 19999
Tenderdash 26656 (though I chose 26262)
HTTPS gRPC Bridge 10000
pshenmic commented 8 months ago

What exactly do you see crashing. I think the right approach should be that it waits until could connect and should not crash. Could you post some logs?

coolaj86 commented 8 months ago

I believe tenderdash is the thing that's the most finicky - erroring out when drive is restarted. It was either that or drive erroring out when tenderdash was restarted.

Next time I see it happen I will post back the logs.

shumkov commented 1 month ago

Tenderdash is failing in some cases when Drive is falling. When Tenderdash is restarting, Drive is not. Please let us know if you see it behaves differently.