visualize_context_switches.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 1,
   "id": "43242b54-7b4c-45e2-9f6c-cd4e154ccdad",
   "metadata": {},
   "outputs": [],
   "source": [
    "import os\n",
    "from distutils.dir_util import copy_tree\n",
    "import itertools\n",
    "import scipy\n",
    "from scipy import stats\n",
    "import pandas as pd\n",
    "import numpy as np\n",
    "import random\n",
    "from datetime import datetime\n",
    "import matplotlib.pyplot as plt\n",
    "import seaborn as sns\n",
    "import pytz\n",
    "from matplotlib.ticker import MultipleLocator, FixedLocator, LogLocator, NullFormatter, ScalarFormatter\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "id": "fb0079c7-3b34-432d-8611-954726cc2ea5",
   "metadata": {},
   "outputs": [],
   "source": [
    "gpu_nodes = {\n",
    "    \"r28n1\", \"r28n2\", \"r28n3\", \"r28n4\", \"r28n5\",\n",
    "    \"r29n1\", \"r29n2\", \"r29n3\", \"r29n4\", \"r29n5\",\n",
    "    \"r30n1\", \"r30n2\", \"r30n3\", \"r30n4\", \"r30n5\", \"r30n6\", \"r30n7\",\n",
    "    \"r31n1\", \"r31n2\", \"r31n3\", \"r31n4\", \"r31n5\", \"r31n6\"\n",
    "    \"r32n1\", \"r32n2\", \"r32n3\", \"r32n4\", \"r32n5\", \"r32n6\", \"r32n7\",\n",
    "    \"r33n2\", \"r33n3\", \"r33n5\", \"r33n6\",\n",
    "    \"r34n1\", \"r34n2\", \"r34n3\", \"r34n4\", \"r34n5\", \"r34n6\", \"r34n7\",\n",
    "    \"r35n1\", \"r35n2\", \"r35n3\", \"r35n4\", \"r35n5\",\n",
    "    \"r36n1\", \"r36n2\", \"r36n3\", \"r36n4\", \"r36n5\",\n",
    "    \"r38n1\", \"r38n2\", \"r38n3\", \"r38n4\", \"r38n5\",\n",
    "}"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "id": "294d2b24-d21c-4e28-b094-b3cbf0c24620",
   "metadata": {},
   "outputs": [],
   "source": [
    "df = pd.read_parquet(\"Q:/Downloads/surf-lisa-machine-data-2019-12-29-23-00-00-2020-08-07-21-59-45.tar/surf-lisa-machine-data-2019-12-29-23-00-00-2020-08-07-21-59-45/processed-surf-dataset/node_context_switches_total\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "id": "49cccc4e-77b4-4dd6-9aa0-4218d061f94b",
   "metadata": {},
   "outputs": [],
   "source": [
    "# We noticed some negative numbers in what should be a cumulative number (see https://docs.splunk.com/observability/gdi/prometheus-node/prometheus-node.html)\n",
    "# We assume these might be node reboots or other events that reset the counter, and thus clip the results to 0.\n",
    "# We assume positive changes are valid.\n",
    "df = df.dropna().diff().clip(lower=0)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "id": "ba90c8b5-4e6f-429c-aec5-8dda79c100ff",
   "metadata": {},
   "outputs": [],
   "source": [
    "visualize_top = 15\n",
    "top_nodes = df.max().sort_values().index[-visualize_top:]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "id": "e1543dc6-7503-4e64-87ae-58eaa4c8a287",
   "metadata": {},
   "outputs": [],
   "source": [
    "def make_plot(ax, series):\n",
    "    ax.plot(series, color=\"lightcoral\")\n",
    "    ax.get_yaxis().set_major_formatter(ScalarFormatter(useMathText=True)) \n",
    "    ax.yaxis.get_offset_text().set_visible(False)\n",
    "    \n",
    "    def get_converted_xticks(ax):\n",
    "        \"\"\"\n",
    "        :param ax:\n",
    "        :return list of day and month strings\n",
    "        \"\"\"\n",
    "        return [pd.to_datetime(int(tick.get_text()), utc=True, unit=\"s\").strftime(\"%d\\n%b\") for tick in ax.get_xticklabels()]\n",
    "    \n",
    "    ax.get_xaxis().set_major_formatter(ScalarFormatter(useMathText=False)) \n",
    "    ax.ticklabel_format(axis='x', style='plain')\n",
    "    ax.xaxis.get_offset_text().set_visible(False)\n",
    "    \n",
    "    fig.tight_layout()\n",
    "    # This call needs to be after the tight_layout call so that the labels have been set!\n",
    "    ax.set_xticklabels(get_converted_xticks(ax))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "id": "60e0ccab-e327-4fd5-af0c-ce00efc8776c",
   "metadata": {},
   "outputs": [],
   "source": [
    "plot_df = df\n",
    "plot_df[\"dt\"] = pd.to_datetime(plot_df.index, utc=True, unit=\"s\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "id": "1709a443-dcd6-4a66-a03f-adb78ec80100",
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "C:\\Users\\laurens\\AppData\\Local\\Temp/ipykernel_12788/3758762637.py:19: UserWarning: FixedFormatter should only be used together with FixedLocator\n",
      "  ax.set_xticklabels(get_converted_xticks(ax))\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 792x360 with 12 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "rows = 3\n",
    "cols = 4\n",
    "fig, axes = plt.subplots(nrows=rows, ncols=cols, sharex=True, sharey=True, figsize=(11, 5))\n",
    "\n",
    "for row in range(rows):\n",
    "    for col in range(cols):\n",
    "        ax = axes[row, col]\n",
    "        node = top_nodes[row*cols + col]\n",
    "        make_plot(ax, plot_df[node])\n",
    "        ax.set_title(node)\n",
    "\n",
    "fig.tight_layout()\n",
    "\n",
    "fig.supxlabel('Date', fontsize=18)\n",
    "fig.supylabel(\"Context Switches ({})\".format(axes[0, 0].get_yaxis().get_major_formatter().get_offset()), fontsize=18)\n",
    "fig.tight_layout()\n",
    "date_time = datetime.now().strftime(\"%Y-%m-%d-%H-%M-%S\")\n",
    "fig.savefig(f\"output/context_switches_top_{rows*cols}_nodes_{date_time}.pdf\", bbox_inches='tight')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "id": "0b740fbb-7032-4fee-8d99-8b9154ab3235",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "r34n1     1590508155\n",
       "r14n3     1591262055\n",
       "r11n3     1591262055\n",
       "r13n13    1591262055\n",
       "r27n22    1591262055\n",
       "r12n29    1591262055\n",
       "r10n20    1591262055\n",
       "r26n11    1591262055\n",
       "r26n13    1591262055\n",
       "r12n18    1591262055\n",
       "r12n32    1596664800\n",
       "r12n19    1596664800\n",
       "dtype: int64"
      ]
     },
     "execution_count": 23,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "plot_df[top_nodes[:12]].idxmax().sort_values()"
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3 (ipykernel)",
   "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": "3.10.0"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 5
}