andy108369

# based on https://github.com/ovrclk/cosmos-omnibus/blob/master/sifchain/deploy.yml
# https://gist.github.com/andy108369/a869d1d43cd8c67c0700c7acebc36838
# 19 March 2023
#
# - 1. upload GPG-encrypted (use `--symmetric` flag) node_key.json & priv_validator_key.json to your Storj or Filebase S3-compatible bucket
# - 2. update REDACTED fields
# - 3. deploy it in Akash Network https://docs.akash.network/guides
---
version: "2.0"

andy108369

---
version: "2.0"

services:
  txgen-app:
    image: busybox
    command:
     - "sh"
     - "-c"
    args:

andy108369

When the user has a certificate generated in the future he gets the following error:

Error while sending manifest to provider. request to https://provider.akash.pro:8443/deployment/7680977/manifest failed, reason: socket hang up

Or

Error while sending manifest to provider. request to https://d3akash.cloud:8443/deployment/7678683/manifest failed, reason: read ECONNRESET

His certificate is valid in the future

andy108369

#!/usr/bin/env bash
# Filename: clean-stale-akash-resources.sh
# Author: andrey.arapov@nixaid.com
# Version: 1.3 - 19 June 2023
# Source: https://gist.github.com/andy108369/f211bf6c06f2a6e3635b20bdfb9f0fca
# Ref. https://docs.akash.network/providers/build-a-cloud-provider/akash-provider-troubleshooting/dangling-deployments

# Make sure you have kubectl, akash and provider-services binaries.
#
# akash https://github.com/ovrclk/akash/releases/latest

andy108369

See the working PostgreSQL deployment SDL manifest for Akash Network in the comments section.

Upgrade PostgreSQL from 13 to 14 in a container

1. Start postgresql:14 container with sleep infinity entrypoint

with the data mounted over /var/lib/postgresql/data directory.

Make sure to either use PGDATA=/var/lib/postgresql/data/<some-other-dir> or rmdir /var/lib/postgresql/data/lost+found when having /var/lib/postgresql/data a mountpoint over a device (say Ceph's RBD), due to initdb not willing to work with the mountpoints (but can be tricked to by removing lost+found directory! :-)).
More here https://www.postgresql.org/message-id/flat/21269e094e35a1b32c426408cc0a4408%40nixaid.com#e4510c5b090c2737e68cfdcf8223020a

andy108369

Final work is here and a working SDL manifest

What's below on this page is outdated:

---

1. Init the DB with postgres:adminpass creds

andy108369

Fix for https://github.com/ovrclk/awesome-akash/blob/b720cd57784f9/chia-madmax/deploy.yaml#L7-L12

    expose:
      - port: 8080
        as: 80
        http_options:
          read_timeout: 3600000
        to:
 - global: true

andy108369

# app-stunnel-client.yml
# https://github.com/ovrclk/stunnel-proxy
---
version: "2.0"

services:
  stunnel-client:
    image: andrey01/stunnel-proxy:v0.0.1
    env:
      - PSK=RHFVUmtrQ0EyVFhVanlmVTJZZXgK

andy108369

# app-stunnel-server.yml
# https://github.com/ovrclk/stunnel-proxy
---
version: "2.0"

services:
  app:
    image: traefik/whoami
    expose:
      - port: 80

andy108369

Enabling Ceph dashboard

---

Make sure it is enabled in the config

Get the admin password: