Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Setup scripts for local dev #30

Closed
wants to merge 1 commit into from
+301 −1
Closed

Setup scripts for local dev #30

Changes from all commits
Commits
Show all changes
1 commit
Select commit
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
28 changes: 28 additions & 0 deletions DEVELOPER_GUIDE.md
View file
Original file line number Diff line number Diff line change
@@ -24,6 +24,34 @@ Ultimately, your directory structure should look like this:
│ └── dashboards-assistant
```

Setup local stack

1. Export keys locally:
```sh
# defaults to us-west-2
export REGION=
export AWS_ACCESS_KEY_ID=
export AWS_SECRET_ACCESS_KEY=
export AWS_SESSION_TOKEN=
```
2. Setup OSD plugins
```sh
# from OSD root folder
cd plugins
# example with security plugin
git clone --depth 1 --branch $VERSION https://github.com/opensearch-project/security-dashboards-plugin.git
cd dashboards-assistant
# NOTE: this will change OSD to be version 2.11.0 for dev purposes
yarn setup
cd ../../
yarn osd bootstrap
```
2. Start local environment
```sh
# from OSD root folder
yarn start:assistant
```

### Build

To build the plugin's distributable zip simply run `yarn build` in the plugin's directory.
3 changes: 2 additions & 1 deletion package.json
View file
Original file line number Diff line number Diff line change
@@ -10,7 +10,8 @@
"plugin-helpers": "node ../../scripts/plugin_helpers",
"prepare": "husky install",
"lint:es": "node ../../scripts/eslint",
"lint": "yarn lint:es"
"lint": "yarn lint:es",
"setup": "scripts/pre_install.sh"
},
"lint-staged": {
"*.{ts,tsx,js,jsx}": [
89 changes: 89 additions & 0 deletions scripts/assistant/add_model.sh
View file
Original file line number Diff line number Diff line change
@@ -0,0 +1,89 @@
#!/usr/bin/env bash
set -e

function add_model() {
[ -z "$ACCESS_KEY_ID" ] && ACCESS_KEY_ID=${AWS_ACCESS_KEY_ID}
[ -z "$SECRET_ACCESS_KEY" ] && SECRET_ACCESS_KEY=${AWS_SECRET_ACCESS_KEY}
[ -z "$SESSION_TOKEN" ] && SESSION_TOKEN=${AWS_SESSION_TOKEN}

ENDPOINT=https://admin:${CREDENTIAL}@${BIND_ADDRESS}:${BIND_PORT}

curl -s -k "${ENDPOINT}/_cluster/settings" -XPUT -H 'Content-Type: application/json' -d '{
"persistent" : {
"plugins.ml_commons.trusted_connector_endpoints_regex":
[ "^https://runtime\\.sagemaker\\..*[a-z0-9-]\\.amazonaws\\.com/.*$",
"^https://api\\.openai\\.com/.*$",
"^https://api\\.cohere\\.ai/.*$",
"^https://bedrock-runtime\\.[a-z0-9-]+\\.amazonaws\\.com/.*$"
]
}
}' | jq

# shellcheck disable=2016
CONNECTOR=$(curl -s -k "${ENDPOINT}/_plugins/_ml/connectors/_create" -XPOST -H 'Content-Type: application/json' -d '{
"name": "BedRock test claude Connector",
"description": "The connector to BedRock service for claude model",
"version": 1,
"protocol": "aws_sigv4",
"parameters": {
"region": "'$REGION'",
"service_name": "bedrock",
"anthropic_version": "bedrock-2023-05-31",
"endpoint": "bedrock.'$REGION'.amazonaws.com",
"auth": "Sig_V4",
"content_type": "application/json"
},
"credential": {
"access_key": "'$ACCESS_KEY_ID'",
"secret_key": "'$SECRET_ACCESS_KEY'",
"session_token": "'$SESSION_TOKEN'"
},
"actions": [
{
"action_type": "predict",
"method": "POST",
"url": "https://bedrock-runtime.'"$REGION"'.amazonaws.com/model/anthropic.claude-v1/invoke",
"headers": {
"content-type": "application/json",
"x-amz-content-sha256": "required"
},
"request_body": "{\"prompt\":\"${parameters.prompt}\", \"max_tokens_to_sample\":${parameters.max_tokens_to_sample}, \"temperature\":${parameters.temperature}, \"anthropic_version\":\"${parameters.anthropic_version}\" }"
}
]
}' | jq -r '.connector_id')
echo "❗connector: ${CONNECTOR}"

GROUP=$(curl -s -k "${ENDPOINT}/_plugins/_ml/model_groups/_register" -XPOST -H 'Content-Type: application/json' -d '{
"name": "test_model_group_public",
"description": "This is a public model group"
}' | jq | grep -oP '(?<=ID: )(.+)(?=\.)|(?<=model_group_id": ")(.+)(?=",)' | head -n 1)
echo "❗group: ${GROUP}"

EMBEDDINGS_TASK=$(curl -s -k "${ENDPOINT}/_plugins/_ml/models/_register?deploy=true" -XPOST -H 'Content-Type: application/json' -d '{
"name": "huggingface/sentence-transformers/all-mpnet-base-v2",
"version": "1.0.1",
"model_group_id": "'$GROUP'",
"model_format": "TORCH_SCRIPT"
}' | jq -r '.task_id')
echo "❗embeddings_task: ${EMBEDDINGS_TASK}"

MODEL=$(curl -s -k "${ENDPOINT}/_plugins/_ml/models/_register?deploy=true" -XPOST -H 'Content-Type: application/json' -d '{
"name": "Claude model on bedrock",
"model_group_id": "'$GROUP'",
"function_name": "remote",
"version": "1.0.0",
"connector_id": "'$CONNECTOR'",
"description": "test model"
}' | jq -r '.model_id')

echo "❗model: ${MODEL}"
sleep 40
EMBEDDINGS_MODEL=$(curl -s -k "${ENDPOINT}/_plugins/_ml/tasks/${EMBEDDINGS_TASK}" | jq -r '.model_id')
echo "❗embeddings_model: ${EMBEDDINGS_MODEL}"

curl -s -k "${ENDPOINT}/.chat-assistant-config/_doc/model-config" -XPOST -H 'Content-Type: application/json' -d '{
"model_type":"claude_bedrock",
"embeddings_model_id":"'$EMBEDDINGS_MODEL'",
"model_id":"'$MODEL'"
}' | jq
}
85 changes: 85 additions & 0 deletions scripts/assistant/start.sh
View file
Original file line number Diff line number Diff line change
@@ -0,0 +1,85 @@
#!/bin/bash

# Copyright OpenSearch Contributors
# SPDX-License-Identifier: Apache-2.0

set -e

. scripts/assistant/utils.sh
. scripts/assistant/add_model.sh

function usage() {
echo ""
echo "This script is used to run OpenSearch Assistant"
echo "--------------------------------------------------------------------------"
echo "Usage: $0 [args]"
echo ""
echo "Optional arguments:"
echo -e "-b BIND_ADDRESS\t, defaults to localhost | 127.0.0.1, can be changed to any IP or domain name for the cluster location."
echo -e "-p BIND_PORT\t, defaults to 5601 depends on OpenSearch or Dashboards, can be changed to any port for the cluster location."
echo -e "-c CREDENTIAL\t(password), defaults to admin"
echo -e "-h\tPrint this message."
echo "--------------------------------------------------------------------------"
}

while getopts ":h:b:p:c:" arg; do
case $arg in
h)
usage
exit 1
;;
b)
BIND_ADDRESS=$OPTARG
;;
p)
BIND_PORT=$OPTARG
;;
c)
CREDENTIAL=$OPTARG
;;
:)
echo "-${OPTARG} requires an argument"
usage
exit 1
;;
?)
echo "Invalid option: -${OPTARG}"
exit 1
;;
esac
done

[ -z "$BIND_ADDRESS" ] && BIND_ADDRESS="localhost"
[ -z "$BIND_PORT" ] && BIND_PORT="9200"
[ -z "$CREDENTIAL" ] && CREDENTIAL="admin"
[ -z "$REGION" ] && REGION="us-west-2"

PARENT_PID_LIST=()

PACKAGE_VERSION=$(yarn --silent pkg-version)

# define assistant path
CWD=$(pwd)
SNAPSHOT_DIR="$CWD/.opensearch"
LOGS_DIR="$SNAPSHOT_DIR/$PACKAGE_VERSION/logs"

# Main function
function execute() {
export initialAdminPassword=$CREDENTIAL
CLUSTER_SETTINGS="snapshot --assistant --security"
CLUSTER_SETTINGS+=" -E plugins.ml_commons.only_run_on_ml_node=true"

run_opensearch || clean
check_opensearch_status
echo "Attempting to add models..."
echo "(Ensure your environment is exporting your credentials)"
(add_model > $LOGS_DIR/add_model.log 2>&1 || clean) &

export OPENSEARCH_USERNAME=kibanaserver
export OPENSEARCH_PASSWORD=kibanaserver
Copy link
Member

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

what if user changed them?

Copy link
Member Author

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Yeah I wish we can change this username for this user since I think it's confusing. This is an internal user. Which I feel like it's not accurate by the description linked as it doesn't feel like a demo user since it has super privileges. It also seems to be able to ignore attempts to remove privileges.

This script utilizes snapshots and anytime that is ran the snapshot will get wiped out clean and will have this super user. Which really is just used for the health check. This part is kind of more like an "easy install". What this does is call to OSD scripts to start up the cluster with security plugin and assistant required plugins. Run all the models stuff and then start up OSD. However, if the user would like to change some stuff then can do that stuff a little bit manually with the following:

In a window run the OpenSearch snapshot:

yarn opensearch snapshot --assistant --security  -E plugins.ml_commons.only_run_on_ml_node=true

run the curl requests.

and then in another window starting up a process of OpenSearch Dashboards

yarn start --security

With that said, I can still modify if a hard blocker. Lemme know!

Copy link
Member

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

i see, i asked because i remember logging into someone's dashboards using kibanaserver and people might want to change that. but since it's for development i think it's ok

$CWD/scripts/use_node $CWD/scripts/opensearch_dashboards --dev --security || clean
}

execute
clean
exit 0
86 changes: 86 additions & 0 deletions scripts/assistant/utils.sh
View file
Original file line number Diff line number Diff line change
@@ -0,0 +1,86 @@
#!/bin/bash

# Copyright OpenSearch Contributors
# SPDX-License-Identifier: Apache-2.0

set -e

function open_artifact() {
artifact_dir=$1
artifact=$2
cd $artifact_dir

# check if artifact provided is URL or attempt if passing by absolute path
if curl -I -L $artifact; then
curl -L $artifact | tar -xz --strip-components=1
else
echo "Artifact is not a URL; attempting to unarchive a local file..."
tar -xf $artifact --strip-components=1
fi
}

# remove the running opensearch process
function clean() {
echo "Attempt to Terminate Process with PID: ${PARENT_PID_LIST[*]}"
for pid_kill in "${PARENT_PID_LIST[@]}"
do
echo "Closing PID $pid_kill"
kill $pid_kill || true
done
PARENT_PID_LIST=()
}

function spawn_process_and_save_PID() {
echo "Spawn '$@'"
eval $@
curr_pid=$!
echo "PID: $curr_pid"
PARENT_PID_LIST+=( $curr_pid )
}

# Print out a textfile line by line
function print_txt() {
while IFS= read -r line; do
echo "text read from $1: $line"
done < $1
}

# this function is used to check the running status of OpenSearch or OpenSearch Dashboards
# $1 is the path to the tmp file which saves the running status
# $2 is the error msg to check
# $3 is the url to curl
# $4 contains arguments that need to be passed to the curl command
function check_status() {
# Calculate end time as 350s from now
check_status_end_time=$(expr 350 + "$(date '+%s')")

while [ ! -e $1 ] || ! grep -q "$2" $1; do
sleep 1
# Stop checking after $check_status_end_time
if [ $check_status_end_time -lt $(date '+%s') ]; then
echo "Error: Status check has timed out"
exit 1
fi
done
}

# Checks the running status of OpenSearch
# it calls check_status and passes the OpenSearch tmp file path, error msg, url, and arguments
# if success, the while loop in the check_status will end and it prints out "OpenSearch is up!"
function check_opensearch_status() {
echo "Checking the status OpenSearch..."
# define other paths and tmp files
OPENSEARCH_FILE='opensearch.log'
OPENSEARCH_LOG_PATH="$LOGS_DIR/$OPENSEARCH_FILE"

OPENSEARCH_MSG="ML configuration initialized successfully"
check_status $OPENSEARCH_LOG_PATH "$OPENSEARCH_MSG" 2>&1
echo "OpenSearch is up!"
}

# Starts OpenSearch
function run_opensearch() {
echo "[ Attempting to start OpenSearch... ]"
rm -rf $LOGS_DIR/opensearch.log
spawn_process_and_save_PID "$CWD/scripts/use_node $CWD/scripts/opensearch $CLUSTER_SETTINGS > $LOGS_DIR/opensearch.log 2>&1 &"
}
11 changes: 11 additions & 0 deletions scripts/pre_install.sh
View file
Original file line number Diff line number Diff line change
@@ -0,0 +1,11 @@
#!/bin/bash

# Copyright OpenSearch Contributors
# SPDX-License-Identifier: Apache-2.0

set -e

[ -z "$SUPPORTED_VERSION" ] && SUPPORTED_VERSION="2.11.0"

cp -r scripts/assistant ../../scripts
sed -i -E "s|(\"version\": \")[^\"]*|\1${SUPPORTED_VERSION}|" ../../package.json