Ladas’s gists

Ladas / mock_metrics.rb

Last active January 17, 2018 14:02

generate mock metrics and labels for containers

	# ActiveRecord::Base.logger = Logger.new(STDOUT)

	NUMBER_OF_CONTAINERS = 20_000 # 100_000
	NUMBER_OF_PROJECTS = 100
	LABELS_PER_CONTAINER_IMAGE = 20
	METRIC_PERIOD = 20.seconds
	METRICS_START = 1.days.ago.utc
	METRICS_END = Time.now.utc

	# Have also archived containers?

Ladas / generate_data_in_parallel.rb

Last active January 10, 2018 15:05

Generates data in parallel, in 8 processe the speed is 86k records per second

	ActiveRecord::Base.logger = Logger.new(STDOUT)

	UNIQUE_INDEXES_FOR_MODELS = {
	# Just having :ems_id & :ems_ref
	ContainerBuild => [:ems_id, :ems_ref],
	ContainerBuildPod => [:ems_id, :ems_ref],
	ContainerGroup => [:ems_id, :ems_ref],
	ContainerLimit => [:ems_id, :ems_ref],
	ContainerNode => [:ems_id, :ems_ref],
	ContainerProject => [:ems_id, :ems_ref],

Ladas / generate_data.rb

Last active January 10, 2018 09:55

Generate huge amount of mock data, can insert around 27.4k records per s into Postgre DB

	ActiveRecord::Base.logger = Logger.new(STDOUT)

	UNIQUE_INDEXES_FOR_MODELS = {
	# Just having :ems_id & :ems_ref
	ContainerBuild => [:ems_id, :ems_ref],
	ContainerBuildPod => [:ems_id, :ems_ref],
	ContainerGroup => [:ems_id, :ems_ref],
	ContainerLimit => [:ems_id, :ems_ref],
	ContainerNode => [:ems_id, :ems_ref],
	ContainerProject => [:ems_id, :ems_ref],

Ladas / perf_capture

Created November 29, 2017 08:53

huntin mem leaks

	rss: 908332 live objects 3829402 gc: 23 pg: 9569 fpg: 0 ep 9569 tp 0---- 2017-11-28 16:19:36 UTC 0
	rss: 1006148 live objects 3900281 gc: 29 pg: 9569 fpg: 0 ep 9569 tp 0---- 2017-11-28 16:22:15 UTC 1
	rss: 1028232 live objects 3886911 gc: 34 pg: 9569 fpg: 0 ep 9569 tp 0---- 2017-11-28 16:25:07 UTC 2
	rss: 1039276 live objects 3885838 gc: 39 pg: 9569 fpg: 0 ep 9569 tp 0---- 2017-11-28 16:27:45 UTC 3
	rss: 1052328 live objects 3894789 gc: 44 pg: 9569 fpg: 0 ep 9569 tp 0---- 2017-11-28 16:30:15 UTC 4
	rss: 1063136 live objects 3878740 gc: 49 pg: 9569 fpg: 0 ep 9569 tp 0---- 2017-11-28 16:32:54 UTC 5
	rss: 1073020 live objects 3890319 gc: 54 pg: 9569 fpg: 0 ep 9569 tp 0---- 2017-11-28 16:35:48 UTC 6
	rss: 1080832 live objects 3895098 gc: 59 pg: 9569 fpg: 0 ep 9569 tp 0---- 2017-11-28 16:38:27 UTC 7
	rss: 1086920 live objects 3900161 gc: 64 pg: 9569 fpg: 0 ep 9569 tp 0---- 2017-11-28 16:41:02 UTC 8
	rss: 1091912 live objects 3885167 gc: 69 pg: 9569 fpg: 0 ep 9569 tp 0---- 2017-11-28 16:43:41 UTC 9

Ladas / queue_tartget_benchmark.rb

Last active October 20, 2017 15:21

	def pod(uid)
	pod = {"kind" => "Pod",
	"apiVersion" => "v1",
	"metadata" =>
	{"name" => "stress6-1-trz0g",
	"generateName" => "stress6-1-",
	"namespace" => "vcr-tests",
	"selfLink" => "/api/v1/namespaces/vcr-tests/pods/stress6-1-trz0g",
	"uid" => uid,
	"resourceVersion" => "1809029",

Ladas / ems_events.rb

Created October 9, 2017 14:36

OSE targeted refresh

	[#<EmsEvent:0x000000000903d000
	id: 3769,
	event_type: "CONTAINER_KILLING",
	message: "Killing container with id docker://deployment:Need to kill Pod",
	timestamp: Mon, 09 Oct 2017 14:15:32 UTC +00:00,
	host_name: nil,
	host_id: nil,
	vm_name: nil,
	vm_location: nil,
	vm_or_template_id: nil,

Ladas / miq.rb

Last active October 6, 2017 09:29

metrics

	{2017-10-05 11:33:14 UTC=>0.0008883925},
	{2017-10-05 11:33:54 UTC=>0.0011175},
	{2017-10-05 11:34:14 UTC=>22.935428365},
	{2017-10-05 11:34:54 UTC=>84.5403323575},
	{2017-10-05 11:35:14 UTC=>88.23617879},
	{2017-10-05 11:35:54 UTC=>55.870726725},
	{2017-10-05 11:36:14 UTC=>80.4185096625},
	{2017-10-05 11:36:54 UTC=>74.556680755},
	{2017-10-05 11:37:14 UTC=>91.34505966},
	{2017-10-05 11:37:54 UTC=>42.44492202},

Ladas / ocp_refresh_mock_data.rb

Created August 9, 2017 14:28

run as "bundle exec rails r ocp_refresh_mock_data.rb"

	require 'manageiq_performance'

	def persister_class
	# ManageIQ::Providers::Kubernetes::Inventory::Persister::ContainerManager
	ManageIQ::Providers::Openshift::Inventory::Persister::ContainerManagerStream # just having :batch saving
	end

	def generate_batches_od_data(ems_name:, total_elements:, batch_size: 1000)
	ems = ExtManagementSystem.find_by(:name => ems_name)
	persister = persister_class.new(

Ladas / ovirt_count_records.rb

Created August 1, 2017 07:04

Count ovirt records, run as: bundle exec rails r ovirt_count_records.rb

	classes = [
	ManageIQ::Providers::Redhat::InfraManager,
	EmsFolder,
	EmsCluster,
	ResourcePool,
	Host,
	Vm,
	Relationship,
	Storage,
	MiqEventDefinition,

Ladas / memusg

Created July 21, 2017 12:02

refresh quick perf test, run as ./memusg bundle exec rails r qe_ocp_refresh_manager_first_refresh.rb <ems_name>, grep evm.log for :save_inventory to get a time

	#!/usr/bin/env bash
	# memusg -- Measure memory usage of processes
	# Usage: memusg COMMAND [ARGS]...
	#
	# Author: Jaeho Shin <netj@sparcs.org>
	# Created: 2010-08-16
	############################################################################
	# Copyright 2010 Jaeho Shin. #
	# #
	# Licensed under the Apache License, Version 2.0 (the "License"); #

Ladislav Smola Ladas