added benchmark

stackql · Dec 27, 2023 · 342fdbe · 342fdbe
1 parent 0c0047d
commit 342fdbe
Show file tree

Hide file tree

Showing 3 changed files with 146 additions and 52 deletions.
diff --git a/notebooks/aws/aws.ipynb b/notebooks/aws/aws.ipynb
@@ -15,7 +15,8 @@
    "metadata": {},
    "outputs": [],
    "source": [
-    "region = 'us-east-1'"
+    "regions = \"'ap-southeast-2', 'us-east-1'\"\n",
+    "region = 'ap-southeast-2'"
    ]
   },
   {
@@ -29,7 +30,7 @@
     "split_part(instanceState, '\\n', 3) as instanceState, \n",
     "count(*) as num_instances\n",
     "FROM aws.ec2.instances \n",
-    "WHERE region = '$region' \n",
+    "WHERE region IN ($regions)\n",
     "GROUP BY split_part(instanceState, '\\n', 3)"
    ]
   },
@@ -49,9 +50,9 @@
    "outputs": [],
    "source": [
     "%%stackql\n",
-    "SELECT instanceType, COUNT(*) as num_instances \n",
+    "SELECT region, instanceType, COUNT(*) as num_instances\n",
     "FROM aws.ec2.instances\n",
-    "WHERE region = '$region'\n",
+    "WHERE region IN ($regions)\n",
     "GROUP BY instanceType"
    ]
   },
@@ -95,10 +96,8 @@
    "metadata": {},
    "outputs": [],
    "source": [
-    "# get multiple regions asynchronously\n",
-    "from pystackql import StackQL\n",
-    "stackql = StackQL(download_dir='/srv/stackql', output='pandas')\n",
-    "stackql.executeStmt(\"REGISTRY PULL aws\")"
+    "%%stackql\n",
+    "select PolicyName, Arn from aws.iam.policies where region = '$region'"
    ]
   },
   {
@@ -107,25 +106,27 @@
    "metadata": {},
    "outputs": [],
    "source": [
-    "# get multiple regions asynchronously\n",
-    "import nest_asyncio, json\n",
-    "nest_asyncio.apply()\n",
-    "import pandas as pd\n",
-    "\n",
-    "regions = [\"ap-southeast-2\", \"us-east-1\"]\n",
-    "\n",
-    "queries = [\n",
-    "    f\"\"\"\n",
-    "    SELECT '{region}' as region, instanceType, COUNT(*) as num_instances\n",
-    "    FROM aws.ec2.instances\n",
-    "    WHERE region = '{region}'\n",
-    "    GROUP BY instanceType\n",
-    "    \"\"\"\n",
-    "    for region in regions\n",
-    "]\n",
-    "\n",
-    "instances_df = await stackql.executeQueriesAsync(queries)\n",
-    "instances_df"
+    "%%stackql\n",
+    "SELECT region, function_name\n",
+    "FROM aws.lambda.functions\n",
+    "WHERE region IN (\n",
+    "'us-east-1',\n",
+    "'us-east-2',\n",
+    "'us-west-1',\n",
+    "'us-west-2',\n",
+    "'ap-south-1',\n",
+    "'ap-northeast-3',\n",
+    "'ap-northeast-2',\n",
+    "'ap-southeast-1',\n",
+    "'ap-southeast-2',\n",
+    "'ap-northeast-1',\n",
+    "'ca-central-1',\n",
+    "'eu-central-1',\n",
+    "'eu-west-1',\n",
+    "'eu-west-2',\n",
+    "'eu-west-3',\n",
+    "'eu-north-1',\n",
+    "'sa-east-1')"
    ]
   },
   {
@@ -134,8 +135,18 @@
    "metadata": {},
    "outputs": [],
    "source": [
-    "%%stackql\n",
-    "select PolicyName, Arn from aws.iam.policies where region = '$region'"
+    "functions_list = _.to_dict(orient='records')"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "from pystackql import StackQL\n",
+    "stackql = StackQL(download_dir='/srv/stackql', app_root='/jupyter/.stackql', output='pandas')\n",
+    "stackql.executeStmt(\"REGISTRY PULL aws\")"
    ]
   },
   {
@@ -144,28 +155,11 @@
    "metadata": {},
    "outputs": [],
    "source": [
-    "import nest_asyncio, json\n",
+    "# get all functions across all regions asynchronously using pystackql\n",
+    "import nest_asyncio\n",
     "nest_asyncio.apply()\n",
-    "import pandas as pd\n",
-    "\n",
-    "regions= [\"us-east-1\",\"us-east-2\",\"us-west-1\",\"us-west-2\",\"ap-south-1\",\"ap-northeast-3\",\"ap-northeast-2\",\"ap-southeast-1\",\n",
-    "         \"ap-southeast-2\",\"ap-northeast-1\",\"ca-central-1\",\"eu-central-1\",\"eu-west-1\",\"eu-west-2\",\"eu-west-3\",\"eu-north-1\",\n",
-    "          \"sa-east-1\"]\n",
-    "\n",
-    "get_fns = [\n",
-    "    f\"\"\"\n",
-    "    SELECT *\n",
-    "    FROM aws.lambda.functions\n",
-    "    WHERE region = '{region}'\n",
-    "    \"\"\"\n",
-    "    for region in regions\n",
-    "]\n",
-    "\n",
-    "functions_df = await stackql.executeQueriesAsync(get_fns)\n",
-    "functions_list = functions_df.to_dict(orient='records')\n",
-    "\n",
     "get_fns_details = [\n",
-    "    f\"\"\"\n",
+    "    f'''\n",
     "    SELECT \n",
     "    function_name,\n",
     "    region,\n",
@@ -177,13 +171,21 @@
     "    FROM aws.lambda.function\n",
     "    WHERE region = '{function['region']}'\n",
     "    AND data__Identifier = '{function['function_name']}'\n",
-    "    \"\"\"\n",
+    "    '''\n",
     "    for function in functions_list\n",
     "]\n",
-    "\n",
     "function_details_df = await stackql.executeQueriesAsync(get_fns_details)\n",
     "function_details_df"
    ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "f"
+   ]
   }
  ],
  "metadata": {

diff --git a/notebooks/aws/benchmarks.ipynb b/notebooks/aws/benchmarks.ipynb
@@ -0,0 +1,92 @@
+{
+ "cells": [
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "from pystackql import StackQL\n",
+    "stackql = StackQL(download_dir='/srv/stackql', execution_concurrency_limit=-1, app_root='/jupyter/.stackql', output='pandas')\n",
+    "stackql.executeStmt(\"REGISTRY PULL aws\")"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "regions = [\n",
+    "'us-east-1',\n",
+    "'us-east-2',\n",
+    "'us-west-1',\n",
+    "'us-west-2',\n",
+    "'ap-south-1',\n",
+    "'ap-northeast-3',\n",
+    "'ap-northeast-2',\n",
+    "'ap-southeast-1',\n",
+    "'ap-southeast-2',\n",
+    "'ap-northeast-1',\n",
+    "'ca-central-1',\n",
+    "'eu-central-1',\n",
+    "'eu-west-1',\n",
+    "'eu-west-2',\n",
+    "'eu-west-3',\n",
+    "'eu-north-1',\n",
+    "'sa-east-1'\n",
+    "]"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "# threading test 1 - stackql native golang threading\n",
+    "\n",
+    "regions_list = \"','\".join(regions)\n",
+    "stackql_query = f\"\"\"\n",
+    "SELECT region, function_name\n",
+    "FROM aws.lambda.functions\n",
+    "WHERE region IN ('{regions_list}')\n",
+    "\"\"\"\n",
+    "\n",
+    "import time\n",
+    "\n",
+    "start_time = time.time()\n",
+    "result = stackql.execute(stackql_query)\n",
+    "end_time = time.time()\n",
+    "\n",
+    "print(f\"Time taken (Go threading): {end_time - start_time} seconds\")"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "# threading test 2 - python threading model\n",
+    "\n",
+    "queries = [f\"SELECT region, function_name FROM aws.lambda.functions WHERE region = '{region}'\" for region in regions]\n",
+    "\n",
+    "import nest_asyncio\n",
+    "nest_asyncio.apply()\n",
+    "\n",
+    "start_time = time.time()\n",
+    "result = await stackql.executeQueriesAsync(queries)\n",
+    "end_time = time.time()\n",
+    "print(f\"Time taken (Python async): {end_time - start_time} seconds\")"
+   ]
+  }
+ ],
+ "metadata": {
+  "language_info": {
+   "name": "python"
+  }
+ },
+ "nbformat": 4,
+ "nbformat_minor": 2
+}
diff --git a/scripts/entrypoint.sh b/scripts/entrypoint.sh
@@ -1,6 +1,6 @@
 #!/bin/sh
 /srv/stackql/stackql --version
 echo "starting stackql server..."
-nohup /srv/stackql/stackql --http.response.pageLimit=-1 --pgsrv.port=5466 srv &
+nohup /srv/stackql/stackql --execution.concurrency.limit=-1 --http.response.pageLimit=-1 --pgsrv.port=5466 srv &
 echo "stackql server started"
 start-notebook.sh --NotebookApp.token=''