Finalize keeper, improve tx confirmation (#76)

* use \_with\_history for confirmation otherwise txs expire after 300
blocks
* Touch up keeper README
* Rename keeper

Dockerfile

@@ -10,13 +10,13 @@ COPY . .
 
 RUN --mount=type=cache,mode=0777,target=/home/root/app/target \
     --mount=type=cache,mode=0777,target=/usr/local/cargo/registry \
-	cargo build --release --bin validator-keeper
+	cargo build --release --bin stakenet-keeper
 
 #########
 
 FROM debian:buster as validator-history
 RUN apt-get update && apt-get install -y ca-certificates
-ENV APP="validator-keeper"
+ENV APP="stakenet-keeper"
 
 COPY --from=builder /usr/src/app/target/release/$APP ./$APP
 

docker-compose.yaml

@@ -1,10 +1,10 @@
 version: "3"
 services:
-  validator-keeper:
+  stakenet-keeper:
     build:
       context: .
       target: validator-history
-    container_name: validator-keeper
+    container_name: stakenet-keeper
     environment:
       - RUST_LOG=${RUST_LOG:-info}
       - SOLANA_METRICS_CONFIG=${SOLANA_METRICS_CONFIG}

keeper-bot-quick-start.md

@@ -15,26 +15,25 @@ solana-keygen new -o ./credentials/keypair.json
 
 ### ENV
 
-In the root directory create a new folder named `config` and create a `.env` file inside of it
+In the root directory create `.env` file
 
 ```bash
-mkdir config
-touch ./config/.env
+touch .env
 ```
 
-Then copy into the `.env` file the contents below. Everything should be set as-is, however consider replacing the `JSON_RPC_URL` and adjusting the `PRIORITY_FEES`
+Then copy into the `.env` file the contents below. Everything should be set as-is, however you will need to include a `JSON_RPC_URL` that can handle getProgramAccounts calls.
 
-```.env
+```bash
 # RPC URL for the cluster
-JSON_RPC_URL=https://api.mainnet-beta.solana.com
+JSON_RPC_URL="INCLUDE YOUR RPC URL HERE"
 
 # Cluster to specify (mainnet, testnet, devnet)
 CLUSTER=mainnet
 
 # Log levels
 RUST_LOG="info,solana_gossip=error,solana_metrics=info"
 
-# Path to keypair used to pay for account creation and execute transactions
+# Path to keypair used to execute tranasactions
 KEYPAIR=./credentials/keypair.json
 
 # Validator history program ID (Pubkey as base58 string)
@@ -46,59 +45,56 @@ TIP_DISTRIBUTION_PROGRAM_ID=4R3gSG8BpU4t19KYj8CfnbtRpnT8gtk4dvTHxVRwc2r7
 # Steward program ID
 STEWARD_PROGRAM_ID=Stewardf95sJbmtcZsyagb2dg4Mo8eVQho8gpECvLx8
 
-# Steward config account
+# Steward config account for JitoSOL
 STEWARD_CONFIG=jitoVjT9jRUyeXHzvCwzPgHj7yWNRhLcUoXtes4wtjv
 
 # Priority Fees in microlamports
 PRIORITY_FEES=20000
 
+# Retry count
+TX_RETRY_COUNT=100
+
+# Confirmation time after submission
+TX_CONFIRMATION_SECONDS=30
+
 # Run flags (true/false)
 RUN_CLUSTER_HISTORY=true
 RUN_COPY_VOTE_ACCOUNTS=true
 RUN_MEV_COMMISSION=true
 RUN_MEV_EARNED=true
 RUN_STEWARD=true
-RUN_EMIT_METRICS=true
-
-################# DEBUGGING AND ORACLE USE ONLY #################
+RUN_EMIT_METRICS=false
 
 # Interval to update Validator History Accounts (in seconds)
-# VALIDATOR_HISTORY_INTERVAL=300
+VALIDATOR_HISTORY_INTERVAL=300
 
 # Interval to run steward (in seconds)
-# STEWARD_INTERVAL=301
+STEWARD_INTERVAL=301
 
 # Interval to emit metrics (in seconds)
-# METRICS_INTERVAL=60
+METRICS_INTERVAL=60
 
-# For ORACLE_AUTHORITY operator
-# RUN_STAKE_UPLOAD=false
-# RUN_GOSSIP_UPLOAD=false
+# For Oracle Authority Only
+RUN_STAKE_UPLOAD=false
+RUN_GOSSIP_UPLOAD=false
 
 # Run with the startup flag set to true
-# FULL_STARTUP=true
+FULL_STARTUP=true
 
 # Running with no_pack set to true skips packing the instructions and will cost more
-# NO_PACK=false
+NO_PACK=false
 
 # Pay for new accounts when necessary
-# PAY_FOR_NEW_ACCOUNTS=false
+PAY_FOR_NEW_ACCOUNTS=false
 
 # Max time in minutes to wait after any fire cycle
-# COOL_DOWN_RANGE=20
+COOL_DOWN_RANGE=20
 
-# Gossip entrypoint in the form of URL:PORT
-# GOSSIP_ENTRYPOINT=
-
-# Metrics upload config
-# For uploading metrics to your private InfluxDB instance
-# SOLANA_METRICS_CONFIG=
-
-# Path to keypair used specifically for submitting permissioned transactions
-# ORACLE_AUTHORITY_KEYPAIR=
+# Metrics upload influx server (optional)
+SOLANA_METRICS_CONFIG=""
 ```
 
-## Running Docker
+## Running Docker image from source
 
 Once the setup is complete use the following commands to run/manage the docker container:
 
@@ -107,39 +103,53 @@ Once the setup is complete use the following commands to run/manage the docker c
 ### Start Docker
 
 ```bash
-docker compose --env-file config/.env up -d --build  validator-keeper --remove-orphans
+docker compose --env-file .env up -d --build  stakenet-keeper --remove-orphans
 ```
 
-### Stop Docker**
+### View Logs
 
 ```bash
-docker stop validator-keeper; docker rm validator-keeper;
+docker logs stakenet-keeper -f
 ```
 
-### View Logs
+### Stop Docker\*\*
+
+```bash
+docker stop stakenet-keeper; docker rm stakenet-keeper;
+```
+
+## Run from Dockerhub
+
+This image is available on Dockerhub at: https://hub.docker.com/r/jitolabs/stakenet-keeper
 
 ```bash
-docker logs validator-keeper -f
+docker pull jitolabs/stakenet-keeper:latest
+docker run -d \
+  --name stakenet-keeper \
+  --env-file .env \
+  -v $(pwd)/credentials:/credentials \
+  --restart on-failure:5 \
+  jitolabs/stakenet-keeper:latest
 ```
 
-## Running Raw
+## Running as Binary
 
 To run the keeper in terminal, build for release and run the program.
 
 ### Build for Release
 
 ```bash
-cargo build --release --bin validator-keeper
+cargo build --release --bin stakenet-keeper
 ```
 
 ### Run Keeper
 
 ```bash
-RUST_LOG=info cargo run --bin validator-keeper --
+RUST_LOG=info ./target/release/stakenet-keeper
 ```
 
 To see all available parameters run:
 
 ```bash
-RUST_LOG=info cargo run --bin validator-keeper -- -h
+RUST_LOG=info ./target/release/stakenet-keeper -h
 ```

keepers/validator-keeper/Cargo.toml → keepers/stakenet-keeper/Cargo.toml

@@ -1,5 +1,5 @@
 [package]
-name = "validator-keeper"
+name = "stakenet-keeper"
 version = "0.1.0"
 edition = "2021"
 description = "Script to keep validator history accounts up to date"

keepers/validator-keeper/README.md → keepers/stakenet-keeper/README.md

@@ -1,53 +1,70 @@
-## Keeper State
+# StakeNet Keeper Bot
+
+## Overview
+
+This is a service that permissionlessly maintains the StakeNet Validator History and Steward programs. It ensures that Validator History data is kept up to date, and that the Steward program is properly moving through its regular operations to maintain the targeted stake pool. Operations, program addresses, and transaction submission details, can be configured via arguments or environment variables.
+
+## Usage
+
+See [keeper-bot-quick-start.md](../../keeper-bot-quick-start.md) for instructions on how to build and run this service.
+
+## Program Layout
+
+### Keeper State
+
 The `KeeperState` keeps track of:
+
 - current epoch data
 - running tally of all operations successes and failures for the given epoch
 - all accounts fetched from the RPC that are needed downstream
 
 Note: All maps are key'd by the `vote_account`
 
-## Program
-
 ### Initialize
-Gather all needed arguments, and initialize the global `KeeperState`. 
+
+Gather all needed arguments, and initialize the global `KeeperState`.
 
 ### Loop
-The forever loop consists of three parts: **Fetch**, **Fire** and **Emit**. There is only ever one **Fetch** and **Emit** section, and there can be several **Fire** sections. 
+
+The forever loop consists of three parts: **Fetch**, **Fire** and **Emit**. There is only ever one **Fetch** and **Emit** section, and there can be several **Fire** sections.
 
 The **Fire** sections can run on independent cadences - say we want the Validator History functions to run every 300sec and we want to emit metrics every 60sec.
 
-The **Fetch** section is run _before_ and **Fire** section. 
-The **Emit** section is *one tick* after any **Fire** section.
+The **Fetch** section is run _before_ and **Fire** section.
+The **Emit** section is _one tick_ after any **Fire** section.
 
 #### Fetch
+
 The **Fetch** section is in charge of three operations:
+
 - Keeping track of the current epoch and resetting the runs and error counts for each operation
 - Creating any missing accounts needed for the downstream **Fires** to run
 - Fetching and updating all of the needed accounts needed downstream
 
 This is accomplished is by running three functions within the **Fetch** section
+
 - `pre_create_update` - Updates epoch, and fetches all needed accounts that are not dependant on any missing accounts.
 - `create_missing_accounts` - Creates the missing accounts, which can be determined by the accounts fetched in the previous step
-- `post_create_update` - Fetches any last accounts that needed the missing accounts 
+- `post_create_update` - Fetches any last accounts that needed the missing accounts
 
 Since this is run before every **FIRE** section, some accounts will be fetched that are not needed. This may seem wasteful but the simplicity of having a synchronized global is worth the cost.
 
-Notes: 
-- The **Fetch** section is the only section that will mutate the `KeeperState`. 
+Notes:
+
+- The **Fetch** section is the only section that will mutate the `KeeperState`.
 - If anything in the **Fetch** section fails, no **Fires** will run
 
 #### Fire
+
 There are several **Fire** sections running at their own cadence. Before any **Fire** section is run, the **Fetch** section will be called.
 
 Each **Fire** is a call to `operations::operation_name::fire` which will fire off the operation and return the new count of runs and errors for that operation to be saved in the `KeeperState`
 
-Notes: 
-- Each **Fire** is self contained, one should not be dependant on another.
-- No **Fire* will fetch any accounts, if there are needs for them, they should be added to the `KeeperState`
+Notes:
 
+- Each **Fire** is self contained, one should not be dependant on another.
+- No \*_Fire_ will fetch any accounts, if there are needs for them, they should be added to the `KeeperState`
 
 #### Emit
-This section emits the state of the Keeper one tick after any operation has been called. This is because we want to emit a failure of any **Fetch** operation, which on failure advances the tick.
-
-
 
+This section emits the state of the Keeper one tick after any operation has been called. This is because we want to emit a failure of any **Fetch** operation, which on failure advances the tick.