Data_Logger.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "from scipy.signal import chirp, sweep_poly\n",
    "import numpy as np\n",
    "import matplotlib.pyplot as plt\n",
    "t = np.linspace(0, 1, 720)\n",
    "container_sim = 20+np.ceil(20*chirp(t, f0=1, f1=15, t1=1, method='linear',phi=179))\n",
    "np.random.seed(seed=1)\n",
    "uzer = np.random.randint(low=1,high=150,size=720)\n",
    "plt.figure(figsize=(10,5), dpi=300)\n",
    "plt.plot(container_sim)\n",
    "plt.show()\n",
    "plt.figure(figsize=(10,5), dpi=300)\n",
    "plt.plot(uzer)\n",
    "plt.show()\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "#### Importing Necessary Libraries\n",
    "import subprocess\n",
    "import time\n",
    "import sys\n",
    "import time\n",
    "import math\n",
    "from kubernetes import client, config\n",
    "import tensorflow as tf\n",
    "import tensorflow.contrib.layers as layers\n",
    "import os\n",
    "os.environ[\"GOOGLE_APPLICATION_CREDENTIALS\"]=\"/home/ubuntu/jupyter/google.json\"\n",
    "output = subprocess.check_output('kubectl get pods', shell=True)\n",
    "tf.reset_default_graph()  # This is to reset the graph\n",
    "\n",
    "############## \n",
    "config.load_kube_config()\n",
    "DEPLOYMENT_NAME = 'wordpress'\n",
    "DEPLOYMENT_NS = 'default'\n",
    "\n",
    "\n",
    "api_instance = client.ExtensionsV1beta1Api()\n",
    "\n",
    "def get_replica_and_ready():\n",
    "    api_response = api_instance.read_namespaced_deployment(DEPLOYMENT_NAME, DEPLOYMENT_NS)\n",
    "    return api_response.status.replicas, api_response.status.ready_replicas\n",
    "\n",
    "def set_replica_num(rnum):\n",
    "    rnum = int(rnum)\n",
    "    if rnum < 1:\n",
    "        rnum = 1\n",
    "    api_response = api_instance.read_namespaced_deployment(DEPLOYMENT_NAME, DEPLOYMENT_NS)\n",
    "    api_response.spec.replicas = rnum\n",
    "    api_instance.patch_namespaced_deployment_scale(DEPLOYMENT_NAME, DEPLOYMENT_NS, api_response)\n",
    "    \n",
    "print(get_replica_and_ready())\n",
    "u_pre = 2\n",
    "set_replica_num(u_pre)\n",
    "\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "#%% Change working directory from the workspace root to the ipynb file location. Turn this addition off with the DataScience.changeDirOnImportExport setting\n",
    "import os\n",
    "try:\n",
    "    os.chdir(os.path.join(os.getcwd(), 'examples'))\n",
    "    print(os.getcwd())\n",
    "except:\n",
    "    pass\n",
    "import pacs_load_tester as load_tester\n",
    "\n",
    "\n",
    "#%%\n",
    "def custom_sensing():\n",
    "    r1,r2 = get_replica_and_ready()\n",
    "    return {'r1':r1 , 'r2':r2}\n",
    "lt = load_tester.pacsLoadTester(hatch_rate=1, temp_stat_max_len=5, base='http://35.232.158.172:8089/')\n",
    "\n",
    "lt.custom_sensing = custom_sensing\n",
    "# Testing the function\n",
    "lt.custom_sensing()\n",
    "from tqdm.auto import tqdm\n",
    "tqdm.pandas()\n",
    "\n",
    "loop_timer = load_tester.TimerClass()\n",
    "total_timer = load_tester.TimerClass()\n",
    "user_sequence = [1,10,20,30,40,50,60,70,80,90,100]\n",
    "container_Request = container_sim\n",
    "lt.custom_sensing = custom_sensing\n",
    "lt.change_count(user_sequence[0])\n",
    "lt.start_capturing()\n",
    "\n",
    "loop_time_in_secs = load_tester.get_loop_time_in_secs('20s')\n",
    "\n",
    "loop_timer.tic()\n",
    "total_timer.tic()\n",
    "results = None\n",
    "\n",
    "for j in container_Request :\n",
    "    for i in tqdm(range(len(container_sim))):\n",
    "        user_count = j\n",
    "        lt.change_count(user_count)\n",
    "\n",
    "        sleep_time = loop_time_in_secs - loop_timer.toc()\n",
    "        if sleep_time > 0:\n",
    "            time.sleep(sleep_time)\n",
    "\n",
    "        loop_timer.tic()\n",
    "\n",
    "        result = lt.get_all_stats()\n",
    "        df_result = pd.DataFrame(data=result)\n",
    "\n",
    "        if results is None:\n",
    "            results = df_result\n",
    "        else:\n",
    "            results = results.append(df_result)\n",
    "\n",
    "\n",
    "\n",
    "        #####Identification\n",
    "        if (result['current_response_time_average'])!=[]:\n",
    "            if result['current_response_time_average'][-1]!=None :\n",
    "                u_pre = container_sim[i]\n",
    "                if np.isnan(u_pre):\n",
    "                    u_pre = default\n",
    "                if u_pre >= 40 :\n",
    "                    u_pre = 40\n",
    "                try:\n",
    "                    set_replica_num(u_pre)\n",
    "                except:\n",
    "                    print(\"Error occured !!!!!\")\n",
    "            print(\"iteration : \",i ,\"average RSPT is: \",result['current_response_time_average'][-1],\"Number of Replica Ready: \",result['custom_r1'][-1], \"Number of Users: \",user_count)\n",
    "lt.stop_test()\n",
    "\n",
    "results, filename = lt.prepare_results_from_df(results)\n",
    "\n",
    "results.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "scrolled": true
   },
   "outputs": [],
   "source": [
    "res = results\n",
    "res = res[res['total_rps'] > 0]\n",
    "responsetime_avg = np.average(res['current_response_time_average'])\n",
    "responsetime_var = np.sqrt(np.var(res['current_response_time_average']))\n",
    "container_avg = np.average(res['custom_r2'])\n",
    "container_var = np.sqrt(np.var(res['custom_r2']))\n",
    "print(\"average res : \", responsetime_avg , \"Var Res : \",responsetime_var, \"average Container : \",container_avg, \"Var Container : \", container_var )\n",
    "import matplotlib.pyplot as plt\n",
    "\n",
    "get_ipython().run_line_magic('matplotlib', 'inline')\n",
    "\n",
    "plt.figure(figsize=(8,18))\n",
    "plt.subplot(511)\n",
    "#plt.plot(res['elapsed_min'], res['min_response_time'], label='min_response_time')\n",
    "#plt.plot(res['elapsed_min'], res['current_response_time_percentile_50'], label='median_response_time')\n",
    "plt.plot(res['elapsed_min'], res['current_response_time_average'], label='avg_response_time')\n",
    "#plt.plot(res['elapsed_min'], res['current_response_time_percentile_95'], label='95th percentile')\n",
    "#plt.plot(res['elapsed_min'], res['max_response_time'], label='max_response_time')\n",
    "\n",
    "plt.xlabel('Time (minutes)')\n",
    "plt.ylabel('Average Response Time (ms)')\n",
    "plt.legend()\n",
    "\n",
    "plt.subplot(512)\n",
    "plt.plot(res['elapsed_min'], res['user_count'])\n",
    "plt.xlabel('Time (minutes)')\n",
    "plt.ylabel('Num of Users')\n",
    "\n",
    "plt.subplot(513)\n",
    "plt.plot(res['elapsed_min'], res['total_rps'])\n",
    "plt.xlabel('Time (minutes)')\n",
    "plt.ylabel('Throughput (req/s)')\n",
    "\n",
    "plt.subplot(514)\n",
    "plt.plot(res['elapsed_min'], res['fail_ratio'])\n",
    "plt.xlabel('Time (minutes)')\n",
    "plt.ylabel('Fail Ratio')\n",
    "\n",
    "plt.subplot(515)\n",
    "plt.plot(res['elapsed_min'], res['custom_r1'])\n",
    "plt.plot(res['elapsed_min'], res['custom_r2'])\n",
    "plt.xlabel('Time (minutes)')\n",
    "plt.ylabel('Custom Value')\n",
    "\n",
    "\n",
    "\n",
    "filename = filename.replace('aa.csv', '')\n",
    "plt.savefig(filename + '.png', dpi=300)\n",
    "plt.savefig(filename + '.pdf')\n",
    "plt.show()\n",
    "\n",
    "\n",
    "lt.stop_test()\n",
    "\n",
    "results, filename = lt.prepare_results_from_df(results)\n",
    "\n",
    "results.head()"
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.6.8"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}