diff --git a/.github/workflows/test.yml b/.github/workflows/test.yml
index 39d5c81064c..230456d698b 100644
--- a/.github/workflows/test.yml
+++ b/.github/workflows/test.yml
@@ -48,7 +48,7 @@ jobs:
     - name: Prepare environment
       run: |
         cp config/database.yml.sample config/database.yml
-        bundle exec rake db:setup
+        bundle exec rake db:create db:migrate
     - name: Tests
       run: bin/rails test:all
     - name: Upload coverage to Codecov
diff --git a/Dockerfile b/Dockerfile
index 5fa668fa76a..ea632a5b47d 100644
--- a/Dockerfile
+++ b/Dockerfile
@@ -43,7 +43,8 @@ RUN \
   zstd-libs \
   linux-headers \
   zlib-dev \
-  tzdata 
+  git \
+  tzdata
 
 WORKDIR /app
 
@@ -79,7 +80,7 @@ RUN bundle config set --local without 'development test assets' && \
 
 
 # Final stage for app image
-FROM base
+FROM base as app
 
 RUN --mount=type=cache,id=dev-apk-cache,sharing=locked,target=/var/cache/apk \
   --mount=type=cache,id=dev-apk-lib,sharing=locked,target=/var/lib/apk \
diff --git a/Gemfile b/Gemfile
index c7cd8f21929..30a27bc387b 100644
--- a/Gemfile
+++ b/Gemfile
@@ -49,6 +49,8 @@ gem "rotp", "~> 6.2"
 gem "unpwn", "~> 1.0"
 gem "webauthn", "~> 3.0"
 
+gem "timescaledb-rails", git: "https://github.com/segiddins/timescaledb-rails", branch: "segiddins/support-for-timescale-in-other-database"
+
 # Admin dashboard
 gem "avo", "~> 2.28"
 gem "pundit", "~> 2.3"
diff --git a/Gemfile.lock b/Gemfile.lock
index cfdc0b3e03d..c05b6b4f86a 100644
--- a/Gemfile.lock
+++ b/Gemfile.lock
@@ -1,3 +1,11 @@
+GIT
+  remote: https://github.com/segiddins/timescaledb-rails
+  revision: fffb8e6d42b953537bf652995a5c27e881938762
+  branch: segiddins/support-for-timescale-in-other-database
+  specs:
+    timescaledb-rails (0.1.5)
+      rails (>= 6.0)
+
 GEM
   remote: https://rubygems.org/
   specs:
@@ -332,7 +340,7 @@ GEM
       timeout
     net-smtp (0.3.3)
       net-protocol
-    nio4r (2.5.8)
+    nio4r (2.5.9)
     nokogiri (1.14.2)
       mini_portile2 (~> 2.8.0)
       racc (~> 1.4)
@@ -685,6 +693,7 @@ DEPENDENCIES
   sprockets-rails (~> 3.4)
   statsd-instrument (~> 3.5)
   terser (~> 1.1)
+  timescaledb-rails!
   toxiproxy (~> 2.0)
   unpwn (~> 1.0)
   validates_formatting_of (~> 0.9)
diff --git a/app/avo/resources/download_resource.rb b/app/avo/resources/download_resource.rb
new file mode 100644
index 00000000000..d9b987e6edf
--- /dev/null
+++ b/app/avo/resources/download_resource.rb
@@ -0,0 +1,30 @@
+class DownloadResource < Avo::BaseResource
+  self.title = :query
+  self.includes = []
+  # self.search_query = -> do
+  #   scope.ransack(id_eq: params[:q], m: "or").result(distinct: false)
+  # end
+
+  self.find_record_method = lambda { |model_class:, id:, params:| # rubocop:disable Lint/UnusedBlockArgument
+    # In case of perfoming action `id` becomes an array of `ids`
+    split = lambda { |s|
+      parts = s.split("_")
+      raise ArgumentError unless parts.size == 4
+      { rubygem_id: parts[0], version_id: parts[1], log_ticket_id: parts[2].presence, occurred_at: parts[3] }
+    }
+
+    if id.is_a?(Array)
+      id.reduce(model_class) { |a, e| a.or.where(split[e]) }
+    else
+      model_class.find_by!(split[id])
+    end
+  }
+
+  # Fields generated from the model
+  field :occurred_at, as: :date_time
+  field :rubygem, as: :belongs_to
+  field :version, as: :belongs_to
+  field :log_ticket, as: :belongs_to
+  field :downloads, as: :number
+  # add fields here
+end
diff --git a/app/controllers/avo/downloads_controller.rb b/app/controllers/avo/downloads_controller.rb
new file mode 100644
index 00000000000..fc3832ea2ce
--- /dev/null
+++ b/app/controllers/avo/downloads_controller.rb
@@ -0,0 +1,4 @@
+# This controller has been generated to enable Rails' resource routes.
+# More information on https://docs.avohq.io/2.0/controllers.html
+class Avo::DownloadsController < Avo::ResourcesController
+end
diff --git a/app/jobs/fastly_log_processor.rb b/app/jobs/fastly_log_processor.rb
index 1fb362fdc58..27e3c01d3d2 100644
--- a/app/jobs/fastly_log_processor.rb
+++ b/app/jobs/fastly_log_processor.rb
@@ -18,13 +18,18 @@ def perform
     log_ticket = LogTicket.pop(key: key, directory: bucket)
     if log_ticket.nil?
       StatsD.increment("fastly_log_processor.extra")
+      Rails.logger.debug { "No log ticket for key=#{key} directory=#{bucket}, ignoring" }
       return
     end
 
-    counts = download_counts(log_ticket)
+    counts_by_bucket = download_counts(log_ticket)
+    counts = counts_by_bucket.transform_values { |h| h.sum { |_, v| v } }
     StatsD.gauge("fastly_log_processor.processed_versions_count", counts.count)
     Rails.logger.info "Processed Fastly log counts: #{counts.inspect}"
 
+    # insert into timescale first since it is idempotent and GemDownload is not
+    record_downloads(counts_by_bucket, log_ticket.id)
+
     processed_count = counts.sum { |_, v| v }
     GemDownload.for_all_gems.with_lock do
       GemDownload.bulk_update(counts)
@@ -54,16 +59,60 @@ def download_counts(log_ticket)
     ok_status           = Rack::Utils::SYMBOL_TO_STATUS_CODE[:ok]
     not_modified_status = Rack::Utils::SYMBOL_TO_STATUS_CODE[:not_modified]
 
-    file.each_line.with_object(Hash.new(0)) do |log_line, accum|
-      path, response_code = log_line.split[10, 2]
+    file.each_line.with_object(Hash.new { |h, k| h[k] = Hash.new(0) }) do |log_line, accum|
+      parts = log_line.split
+      path, response_code = parts[10, 2]
       case response_code.to_i
       # Only count successful downloads
       # NB: we consider a 304 response a download attempt
       when ok_status, not_modified_status
         if (match = PATH_PATTERN.match(path))
-          accum[match[:path]] += 1
+          timestamp = parts[0].sub!(/\A<\d+>/, "")
+          accum[match[:path]][truncated_timestamp(timestamp)] += 1
         end
       end
     end
   end
+
+  def truncated_timestamp(timestamp)
+    return unless timestamp
+    time = Time.iso8601(timestamp)
+    time.change(min: time.min / 15 * 15)
+  rescue Date::Error
+    nil
+  end
+
+  VERSION_PLUCK_ID_LIMIT = 1000
+  DOWNLOAD_UPSERT_LIMIT = 1000
+
+  def record_downloads(counts_by_bucket, log_ticket_id)
+    ids_by_full_name = counts_by_bucket.each_key.each_slice(VERSION_PLUCK_ID_LIMIT).with_object({}) do |full_names, hash|
+      Version.where(full_name: full_names).pluck(:full_name, :id, :rubygem_id).each { |full_name, *ids| hash[full_name] = ids }
+    end
+
+    counts_by_bucket
+      .lazy
+      .flat_map do |path, buckets|
+        versions = ids_by_full_name[path]
+        if versions.nil?
+          Rails.logger.debug { "No version found for path=#{path}" }
+          next
+        end
+
+        version_id, rubygem_id = versions
+
+        buckets.map do |occurred_at, downloads|
+          { version_id:, rubygem_id:, occurred_at:, downloads:, log_ticket_id: }
+        end
+      end # rubocop:disable Style/MultilineBlockChain
+      .compact
+      .each_slice(DOWNLOAD_UPSERT_LIMIT) do |inserts|
+        # next unless Download.hypertable?
+        Download.upsert_all(
+          inserts,
+          update_only: %i[downloads],
+          unique_by: %i[rubygem_id version_id occurred_at log_ticket_id]
+        )
+      end
+  end
 end
diff --git a/app/models/download.rb b/app/models/download.rb
new file mode 100644
index 00000000000..f29c054c112
--- /dev/null
+++ b/app/models/download.rb
@@ -0,0 +1,40 @@
+class Download < DownloadRecord
+  def query = { rubygem_id:, version_id:, log_ticket_id:, occurred_at: occurred_at.iso8601 }
+  def id = query.values.join("_")
+
+  belongs_to :rubygem, validate: false, optional: true
+  belongs_to :version, validate: false, optional: true
+  belongs_to :log_ticket, validate: false, optional: true
+
+  validates :occurred_at, presence: true
+  validates :downloads, presence: true
+
+  def self.suffix = nil
+  def self.time_period = nil
+
+  def self.class_name_for(suffix:, time_period:)
+    raise ArgumentError if suffix && !time_period
+    [time_period.iso8601, suffix].compact.join("_").classify
+  end
+
+  [15.minutes, 1.day, 1.month, 1.year].each do |time_period|
+    (%w[all_versions all_gems] << nil).each do |suffix|
+      table_name = "#{Download.table_name}_#{time_period.inspect.parameterize(separator: '_')}#{"_#{suffix}" if suffix}"
+      ::Download.class_eval(<<~RUBY, __FILE__, __LINE__ + 1) # rubocop:disable Style/DocumentDynamicEvalDefinition
+        class #{class_name_for(suffix:, time_period:)} < Download
+          attribute :downloads, :integer
+
+          def readonly?
+            true
+          end
+
+          def self.table_name = #{table_name.dump}
+
+          def self.suffix = #{suffix&.dump || :nil}
+
+          def self.time_period = #{time_period.inspect.parameterize(separator: '_').dump}
+        end
+      RUBY
+    end
+  end
+end
diff --git a/app/models/download_record.rb b/app/models/download_record.rb
new file mode 100644
index 00000000000..bc168e13318
--- /dev/null
+++ b/app/models/download_record.rb
@@ -0,0 +1,7 @@
+class DownloadRecord < ApplicationRecord
+  include Timescaledb::Rails::Model
+
+  self.abstract_class = true
+
+  connects_to database: { writing: :downloads, reading: :downloads }
+end
diff --git a/app/policies/download_policy.rb b/app/policies/download_policy.rb
new file mode 100644
index 00000000000..67f027363e6
--- /dev/null
+++ b/app/policies/download_policy.rb
@@ -0,0 +1,15 @@
+class DownloadPolicy < ApplicationPolicy
+  class Scope < Scope
+    def resolve
+      scope.all
+    end
+  end
+
+  def avo_index?
+    true
+  end
+
+  def avo_show?
+    true
+  end
+end
diff --git a/config/brakeman.ignore b/config/brakeman.ignore
index fa1ea084191..e934ed6ecd6 100644
--- a/config/brakeman.ignore
+++ b/config/brakeman.ignore
@@ -1,5 +1,29 @@
 {
   "ignored_warnings": [
+    {
+      "warning_type": "Dangerous Eval",
+      "warning_code": 13,
+      "fingerprint": "8c3ecf6e4ea1f5fdd60b0cf1ea5250b776da54c6083b88fa6b06ace80b667bac",
+      "check_name": "Evaluation",
+      "message": "User input in eval",
+      "file": "app/models/download.rb",
+      "line": 31,
+      "link": "https://brakemanscanner.org/docs/warning_types/dangerous_eval/",
+      "code": "::Download.class_eval(\"class #{class_name_for(:suffix => suffix, :time_period => time_period)} < Download\\n  attribute :downloads, :integer\\n\\n  def readonly?\\n    true\\n  end\\n\\n  def self.table_name = #{\"#{Download.table_name}_#{time_period.inspect.parameterize(:separator => \"_\")}#{\"_#{suffix}\" if suffix}\".dump}\\n\\n  def self.suffix = #{(suffix.dump or :nil)}\\n\\n  def self.time_period = #{time_period.inspect.parameterize(:separator => \"_\").dump}\\nend\\n\", \"app/models/download.rb\", 24)",
+      "render_path": null,
+      "location": {
+        "type": "method",
+        "class": "Download",
+        "method": null
+      },
+      "user_input": "Download.table_name",
+      "confidence": "High",
+      "cwe_id": [
+        913,
+        95
+      ],
+      "note": "Dyanmically creating a matrix of subclasses"
+    },
     {
       "warning_type": "Mass Assignment",
       "warning_code": 105,
@@ -7,7 +31,7 @@
       "check_name": "PermitAttributes",
       "message": "Potentially dangerous key allowed for mass assignment",
       "file": "app/controllers/api/v1/hook_relay_controller.rb",
-      "line": 23,
+      "line": 19,
       "link": "https://brakemanscanner.org/docs/warning_types/mass_assignment/",
       "code": "params.permit(:attempts, :account_id, :hook_id, :id, :max_attempts, :status, :stream, :failure_reason, :completed_at, :created_at, :request => ([:target_url]))",
       "render_path": null,
@@ -24,6 +48,6 @@
       "note": "account_id is used to validate that the request indeed comes from hook relay"
     }
   ],
-  "updated": "2023-03-01 02:46:07 -0800",
+  "updated": "2023-04-02 21:48:50 -0700",
   "brakeman_version": "5.4.1"
 }
diff --git a/config/database.yml.sample b/config/database.yml.sample
index bfc241051bd..505e8dadf61 100644
--- a/config/database.yml.sample
+++ b/config/database.yml.sample
@@ -3,34 +3,69 @@ default: &default
   encoding: utf8
   username: postgres
 
-development:
-  <<: *default
-  database: rubygems_development
-  host: localhost
-  password: devpassword
-  pool: 5
-  timeout: 5000
+timescale: &timescale
+  adapter: postgresql
+  encoding: utf8
+  username: postgres
+  migrations_paths: db/downloads_migrate
+  port: 5434
 
+development:
+  primary:
+    <<: *default
+    database: rubygems_development
+    host: localhost
+    password: devpassword
+    pool: 5
+    timeout: 5000
+  downloads:
+    <<: *timescale
+    database: rubygems_tsdb_development
+    host: localhost
+    pool: 5
+    timeout: 5000
 
 test:
-  <<: *default
-  database: rubygems_test
-  host: localhost
-  min_messages: warning
-  password: testpassword
-  pool: 5
-  timeout: 5000
+  primary:
+    <<: *default
+    database: rubygems_test
+    host: localhost
+    min_messages: warning
+    password: testpassword
+    pool: 5
+    timeout: 5000
+  downloads:
+    <<: *timescale
+    database: rubygems_tsdb_test
+    host: localhost
+    min_messages: warning
+    pool: 5
+    timeout: 5000
 
 staging:
-  <<: *default
-  database: rubygems_staging
-  min_messages: error
-  pool: 30
-  reconnect: true
+  primary:
+    <<: *default
+    database: rubygems_staging
+    min_messages: error
+    pool: 30
+    reconnect: true
+  downloads:
+    <<: *timescale
+    database: rubygems_tsdb_staging
+    min_messages: error
+    pool: 30
+    reconnect: true
 
 production:
-  <<: *default
-  database: rubygems_production
-  min_messages: error
-  pool: 30
-  reconnect: true
+  primary:
+    <<: *default
+    database: rubygems_production
+    min_messages: error
+    pool: 30
+    reconnect: true
+  downloads:
+    <<: *timescale
+    database: rubygems_tsdb_production
+    min_messages: error
+    pool: 30
+    reconnect: true
diff --git a/config/initializers/timescale.rb b/config/initializers/timescale.rb
new file mode 100644
index 00000000000..9d0e431b078
--- /dev/null
+++ b/config/initializers/timescale.rb
@@ -0,0 +1,5 @@
+Rails.application.configure do
+  config.after_initialize do
+    Timescaledb::Rails::ApplicationRecord.connects_to database: { writing: :downloads, reading: :downloads }
+  end
+end
diff --git a/db/downloads_migrate/20230308211125_create_downloads.rb b/db/downloads_migrate/20230308211125_create_downloads.rb
new file mode 100644
index 00000000000..4e79c2618b9
--- /dev/null
+++ b/db/downloads_migrate/20230308211125_create_downloads.rb
@@ -0,0 +1,14 @@
+class CreateDownloads < ActiveRecord::Migration[7.0]
+  def change
+    create_table :downloads, id: false do |t|
+      t.integer :rubygem_id, null: false
+      t.integer :version_id, null: false
+      t.integer :downloads, null: false
+      t.integer :log_ticket_id, null: true
+      t.timestamptz :occurred_at, null: false
+    end
+
+    add_index :downloads, [:rubygem_id, :version_id, :occurred_at, :log_ticket_id], unique: true, name: 'idx_downloads_by_version_log_ticket'
+    create_hypertable :downloads, :occurred_at, chunk_time_interval: '7 days', partitioning_column: 'rubygem_id'
+  end
+end
diff --git a/db/downloads_migrate/20230308224729_create_downloads_continuous_aggregates.rb b/db/downloads_migrate/20230308224729_create_downloads_continuous_aggregates.rb
new file mode 100644
index 00000000000..321e88b9790
--- /dev/null
+++ b/db/downloads_migrate/20230308224729_create_downloads_continuous_aggregates.rb
@@ -0,0 +1,101 @@
+class CreateDownloadsContinuousAggregates < ActiveRecord::Migration[7.0]
+  disable_ddl_transaction!
+
+  def quote_table_name(...)
+    connection.quote_table_name(...)
+  end
+
+  def continuous_aggregate(
+    duration:,
+    from: Download.table_name,
+    start_offset:,
+    end_offset:,
+    schedule_window:,
+    retention: start_offset
+  )
+    name = "downloads_" + duration.inspect.parameterize(separator: '_')
+
+    create_continuous_aggregate(
+      name,
+      Download.
+        time_bucket(duration.iso8601, quote_table_name("#{from}.occurred_at"), select_alias: 'occurred_at').
+        group(:rubygem_id, :version_id).
+        select(:rubygem_id, :version_id).
+        sum(quote_table_name("#{from}.downloads"), :downloads).
+        reorder(nil).
+        from(from).
+        to_sql,
+      force: true
+    )
+    add_continuous_aggregate_policy(name, start_offset&.iso8601, end_offset&.iso8601, schedule_window)
+    add_hypertable_retention_policy(name, retention.iso8601) if retention
+
+    all_versions_name = name + "_all_versions"
+    create_continuous_aggregate(
+      all_versions_name,
+      Download.
+        time_bucket(duration.iso8601, quote_table_name("#{name}.occurred_at"), select_alias: 'occurred_at').
+        group(:rubygem_id).
+        select(:rubygem_id).
+        sum(quote_table_name("#{name}.downloads"), :downloads).
+        from(name).
+        to_sql,
+        force: true
+    )
+    add_continuous_aggregate_policy(all_versions_name, start_offset&.iso8601, end_offset&.iso8601, schedule_window)
+    add_hypertable_retention_policy(all_versions_name, retention.iso8601) if retention
+
+    all_gems_name = name + "_all_gems"
+    create_continuous_aggregate(
+      all_gems_name,
+      Download.
+        time_bucket(duration.iso8601, quote_table_name("#{all_versions_name}.occurred_at"), select_alias: 'occurred_at').
+        sum(quote_table_name("#{all_versions_name}.downloads"), :downloads).
+        from(all_versions_name).
+        to_sql,
+        force: true
+    )
+    add_continuous_aggregate_policy(all_gems_name, start_offset&.iso8601, end_offset&.iso8601, schedule_window)
+    add_hypertable_retention_policy(all_gems_name, retention.iso8601) if retention
+
+    name
+  end
+
+  def change
+    # https://github.com/timescale/timescaledb/issues/5474
+    Download.create!(version_id: 0, rubygem_id: 0, occurred_at: Time.at(0), downloads: 0)
+
+    from = continuous_aggregate(
+      duration: 15.minutes,
+      start_offset: 1.week,
+      end_offset: 1.hour,
+      schedule_window: 1.hour
+    )
+
+    from = continuous_aggregate(
+      duration: 1.day,
+      start_offset: 2.years,
+      end_offset: 1.day,
+      schedule_window: 12.hours,
+      from:
+    )
+
+    from = continuous_aggregate(
+      duration: 1.month,
+      start_offset: nil,
+      end_offset: 1.month,
+      schedule_window: 1.day,
+      retention: nil,
+      from:
+    )
+
+    from = continuous_aggregate(
+      duration: 1.year,
+      start_offset: nil,
+      end_offset: 1.year,
+      schedule_window: 1.month,
+      retention: nil,
+      from:
+    )
+  end
+end
diff --git a/db/downloads_schema.rb b/db/downloads_schema.rb
new file mode 100644
index 00000000000..eba23de4f38
--- /dev/null
+++ b/db/downloads_schema.rb
@@ -0,0 +1,159 @@
+# This file is auto-generated from the current state of the database. Instead
+# of editing this file, please use the migrations feature of Active Record to
+# incrementally modify your database, and then regenerate this schema definition.
+#
+# This file is the source Rails uses to define your schema when running `bin/rails
+# db:schema:load`. When creating a new database, `bin/rails db:schema:load` tends to
+# be faster and is potentially less error prone than running all of your
+# migrations from scratch. Old migrations may fail to apply correctly if those
+# migrations use external dependencies or application code.
+#
+# It's strongly recommended that you check this file into your version control system.
+
+ActiveRecord::Schema[7.0].define(version: 2023_03_08_224729) do
+  # These are extensions that must be enabled in order to support this database
+  enable_extension "plpgsql"
+  enable_extension "timescaledb"
+  enable_extension "timescaledb_toolkit"
+
+  create_table "downloads", id: false, force: :cascade do |t|
+    t.integer "rubygem_id", null: false
+    t.integer "version_id", null: false
+    t.integer "downloads", null: false
+    t.integer "log_ticket_id"
+    t.timestamptz "occurred_at", null: false
+    t.index ["occurred_at"], name: "downloads_occurred_at_idx", order: :desc
+    t.index ["rubygem_id", "version_id", "occurred_at", "log_ticket_id"], name: "idx_downloads_by_version_log_ticket", unique: true
+  end
+
+  create_hypertable "downloads", "occurred_at", chunk_time_interval: "7 days"
+
+  create_continuous_aggregate "downloads_15_minutes", <<-SQL
+    SELECT time_bucket('PT15M'::interval, downloads.occurred_at) AS occurred_at,
+      downloads.rubygem_id,
+      downloads.version_id,
+      sum(downloads.downloads) AS downloads
+     FROM downloads
+    GROUP BY (time_bucket('PT15M'::interval, downloads.occurred_at)), downloads.rubygem_id, downloads.version_id;
+  SQL
+
+  add_continuous_aggregate_policy "downloads_15_minutes", "1 week", "1 hour", "1 hour"
+
+  create_continuous_aggregate "downloads_1_day", <<-SQL
+    SELECT time_bucket('P1D'::interval, downloads_15_minutes.occurred_at) AS occurred_at,
+      downloads_15_minutes.rubygem_id,
+      downloads_15_minutes.version_id,
+      sum(downloads_15_minutes.downloads) AS downloads
+     FROM downloads_15_minutes
+    GROUP BY (time_bucket('P1D'::interval, downloads_15_minutes.occurred_at)), downloads_15_minutes.rubygem_id, downloads_15_minutes.version_id;
+  SQL
+
+  add_continuous_aggregate_policy "downloads_1_day", "2 years", "1 day", "12 hours"
+
+  create_continuous_aggregate "downloads_1_month", <<-SQL
+    SELECT time_bucket('P1M'::interval, downloads_1_day.occurred_at) AS occurred_at,
+      downloads_1_day.rubygem_id,
+      downloads_1_day.version_id,
+      sum(downloads_1_day.downloads) AS downloads
+     FROM downloads_1_day
+    GROUP BY (time_bucket('P1M'::interval, downloads_1_day.occurred_at)), downloads_1_day.rubygem_id, downloads_1_day.version_id;
+  SQL
+
+  add_continuous_aggregate_policy "downloads_1_month", nil, "1 month", "1 day"
+
+  create_continuous_aggregate "downloads_1_year", <<-SQL
+    SELECT time_bucket('P1Y'::interval, downloads_1_month.occurred_at) AS occurred_at,
+      downloads_1_month.rubygem_id,
+      downloads_1_month.version_id,
+      sum(downloads_1_month.downloads) AS downloads
+     FROM downloads_1_month
+    GROUP BY (time_bucket('P1Y'::interval, downloads_1_month.occurred_at)), downloads_1_month.rubygem_id, downloads_1_month.version_id;
+  SQL
+
+  add_continuous_aggregate_policy "downloads_1_year", nil, "1 year", "1 month"
+
+  create_continuous_aggregate "downloads_1_year_all_versions", <<-SQL
+    SELECT time_bucket('P1Y'::interval, downloads_1_year.occurred_at) AS occurred_at,
+      downloads_1_year.rubygem_id,
+      sum(downloads_1_year.downloads) AS downloads
+     FROM downloads_1_year
+    GROUP BY (time_bucket('P1Y'::interval, downloads_1_year.occurred_at)), downloads_1_year.rubygem_id
+    ORDER BY (time_bucket('P1Y'::interval, downloads_1_year.occurred_at));
+  SQL
+
+  add_continuous_aggregate_policy "downloads_1_year_all_versions", nil, "1 year", "1 month"
+
+  create_continuous_aggregate "downloads_1_year_all_gems", <<-SQL
+    SELECT time_bucket('P1Y'::interval, downloads_1_year_all_versions.occurred_at) AS occurred_at,
+      sum(downloads_1_year_all_versions.downloads) AS downloads
+     FROM downloads_1_year_all_versions
+    GROUP BY (time_bucket('P1Y'::interval, downloads_1_year_all_versions.occurred_at))
+    ORDER BY (time_bucket('P1Y'::interval, downloads_1_year_all_versions.occurred_at));
+  SQL
+
+  add_continuous_aggregate_policy "downloads_1_year_all_gems", nil, "1 year", "1 month"
+
+  create_continuous_aggregate "downloads_15_minutes_all_versions", <<-SQL
+    SELECT time_bucket('PT15M'::interval, downloads_15_minutes.occurred_at) AS occurred_at,
+      downloads_15_minutes.rubygem_id,
+      sum(downloads_15_minutes.downloads) AS downloads
+     FROM downloads_15_minutes
+    GROUP BY (time_bucket('PT15M'::interval, downloads_15_minutes.occurred_at)), downloads_15_minutes.rubygem_id
+    ORDER BY (time_bucket('PT15M'::interval, downloads_15_minutes.occurred_at));
+  SQL
+
+  add_continuous_aggregate_policy "downloads_15_minutes_all_versions", "1 week", "1 hour", "1 hour"
+
+  create_continuous_aggregate "downloads_15_minutes_all_gems", <<-SQL
+    SELECT time_bucket('PT15M'::interval, downloads_15_minutes_all_versions.occurred_at) AS occurred_at,
+      sum(downloads_15_minutes_all_versions.downloads) AS downloads
+     FROM downloads_15_minutes_all_versions
+    GROUP BY (time_bucket('PT15M'::interval, downloads_15_minutes_all_versions.occurred_at))
+    ORDER BY (time_bucket('PT15M'::interval, downloads_15_minutes_all_versions.occurred_at));
+  SQL
+
+  add_continuous_aggregate_policy "downloads_15_minutes_all_gems", "1 week", "1 hour", "1 hour"
+
+  create_continuous_aggregate "downloads_1_day_all_versions", <<-SQL
+    SELECT time_bucket('P1D'::interval, downloads_1_day.occurred_at) AS occurred_at,
+      downloads_1_day.rubygem_id,
+      sum(downloads_1_day.downloads) AS downloads
+     FROM downloads_1_day
+    GROUP BY (time_bucket('P1D'::interval, downloads_1_day.occurred_at)), downloads_1_day.rubygem_id
+    ORDER BY (time_bucket('P1D'::interval, downloads_1_day.occurred_at));
+  SQL
+
+  add_continuous_aggregate_policy "downloads_1_day_all_versions", "2 years", "1 day", "12 hours"
+
+  create_continuous_aggregate "downloads_1_day_all_gems", <<-SQL
+    SELECT time_bucket('P1D'::interval, downloads_1_day_all_versions.occurred_at) AS occurred_at,
+      sum(downloads_1_day_all_versions.downloads) AS downloads
+     FROM downloads_1_day_all_versions
+    GROUP BY (time_bucket('P1D'::interval, downloads_1_day_all_versions.occurred_at))
+    ORDER BY (time_bucket('P1D'::interval, downloads_1_day_all_versions.occurred_at));
+  SQL
+
+  add_continuous_aggregate_policy "downloads_1_day_all_gems", "2 years", "1 day", "12 hours"
+
+  create_continuous_aggregate "downloads_1_month_all_versions", <<-SQL
+    SELECT time_bucket('P1M'::interval, downloads_1_month.occurred_at) AS occurred_at,
+      downloads_1_month.rubygem_id,
+      sum(downloads_1_month.downloads) AS downloads
+     FROM downloads_1_month
+    GROUP BY (time_bucket('P1M'::interval, downloads_1_month.occurred_at)), downloads_1_month.rubygem_id
+    ORDER BY (time_bucket('P1M'::interval, downloads_1_month.occurred_at));
+  SQL
+
+  add_continuous_aggregate_policy "downloads_1_month_all_versions", nil, "1 month", "1 day"
+
+  create_continuous_aggregate "downloads_1_month_all_gems", <<-SQL
+    SELECT time_bucket('P1M'::interval, downloads_1_month_all_versions.occurred_at) AS occurred_at,
+      sum(downloads_1_month_all_versions.downloads) AS downloads
+     FROM downloads_1_month_all_versions
+    GROUP BY (time_bucket('P1M'::interval, downloads_1_month_all_versions.occurred_at))
+    ORDER BY (time_bucket('P1M'::interval, downloads_1_month_all_versions.occurred_at));
+  SQL
+
+  add_continuous_aggregate_policy "downloads_1_month_all_gems", nil, "1 month", "1 day"
+
+end
diff --git a/docker-compose.yml b/docker-compose.yml
index 065a162e523..67079143f6e 100644
--- a/docker-compose.yml
+++ b/docker-compose.yml
@@ -6,6 +6,12 @@ services:
       - "5432:5432"
     environment:
       - POSTGRES_HOST_AUTH_METHOD=trust
+  timescale:
+    image: timescale/timescaledb-ha:pg15-ts2.10-latest
+    ports:
+      - "5434:5432"
+    environment:
+      - POSTGRES_HOST_AUTH_METHOD=trust
   cache:
     image: memcached:1.4.24
     ports:
diff --git a/script/build_docker.sh b/script/build_docker.sh
index 9fddfc14ce5..384ecb28a30 100755
--- a/script/build_docker.sh
+++ b/script/build_docker.sh
@@ -24,10 +24,10 @@ docker buildx build --cache-from=type=local,src=/tmp/.buildx-cache \
   --build-arg REVISION="$GITHUB_SHA" \
   .
 
-docker run -e RAILS_ENV=production -e SECRET_KEY_BASE=1234 -e DATABASE_URL=postgresql://localhost \
+docker run -e RAILS_ENV=production -e SECRET_KEY_BASE=1234 -e DATABASE_URL=postgresql://localhost -e DOWNLOADS_DATABASE_URL=postgresql://localhost:5434 \
   --net host "$DOCKER_TAG" \
   -- bin/rails db:create db:migrate
-docker run -d -e RAILS_ENV=production -e SECRET_KEY_BASE=1234 -e DATABASE_URL=postgresql://localhost \
+docker run -d -e RAILS_ENV=production -e SECRET_KEY_BASE=1234 -e DATABASE_URL=postgresql://localhost -e DOWNLOADS_DATABASE_URL=postgresql://localhost:5434 \
   --net host "$DOCKER_TAG" \
   -- puma --environment production --config /app/config/puma.rb
 
diff --git a/test/factories.rb b/test/factories.rb
index a878f9d04b7..570dc3ee698 100644
--- a/test/factories.rb
+++ b/test/factories.rb
@@ -1,4 +1,12 @@
 FactoryBot.define do
+  factory :download do
+    rubygem
+    version
+    downloads { 1 }
+    log_ticket
+    occurred_at { "2023-03-08 13:11:25" }
+  end
+
   factory :admin_github_user, class: "Admin::GitHubUser" do
     login { "jackson-keeling" }
     avatar_url { "MyString" }
diff --git a/test/integration/avo/downloads_controller_test.rb b/test/integration/avo/downloads_controller_test.rb
new file mode 100644
index 00000000000..6059f65c7fa
--- /dev/null
+++ b/test/integration/avo/downloads_controller_test.rb
@@ -0,0 +1,26 @@
+require "test_helper"
+
+class Avo::GemDownloadsControllerTest < ActionDispatch::IntegrationTest
+  include AdminHelpers
+
+  test "getting downloads as admin" do
+    admin_sign_in_as create(:admin_github_user, :is_admin)
+
+    get avo.resources_downloads_path
+
+    assert_response :success
+
+    version = create(:version)
+    download = create(:download, version: version)
+
+    get avo.resources_downloads_path
+
+    assert_response :success
+    page.assert_text download.downloads.to_s
+
+    get avo.resources_download_path(download)
+
+    assert_response :success
+    page.assert_text download.downloads.to_s
+  end
+end
diff --git a/test/jobs/fastly_log_processor_job_test.rb b/test/jobs/fastly_log_processor_job_test.rb
index ed5d46b048f..843cc475013 100644
--- a/test/jobs/fastly_log_processor_job_test.rb
+++ b/test/jobs/fastly_log_processor_job_test.rb
@@ -7,11 +7,24 @@ class FastlyLogProcessorJobTest < ActiveJob::TestCase
     @sample_log = Rails.root.join("test", "sample_logs", "fastly-fake.log").read
 
     @sample_log_counts = {
-      "bundler-1.10.6" => 2,
-      "json-1.8.3-java" => 2,
-      "json-1.8.3" => 1,
-      "json-1.8.2" => 4,
-      "no-such-gem-1.2.3" => 1
+      "bundler-1.10.6" => {
+        Time.utc(2015, 11, 30, 21, 0, 0) => 2
+      },
+      "json-1.8.3-java" => {
+        Time.utc(2015, 11, 30, 21, 0, 0) => 2
+      },
+      "json-1.8.3" => {
+        Time.utc(2015, 11, 30, 21, 0, 0) => 1
+      },
+      "json-1.8.2" => {
+        Time.utc(2015, 11, 29, 21, 0, 0) => 1,
+        Time.utc(2015, 11, 30, 21, 0, 0) => 1,
+        Time.utc(2015, 11, 30, 21, 30, 0) => 1,
+        Time.utc(2015, 12, 30, 21, 0, 0) => 1
+      },
+      "no-such-gem-1.2.3" => {
+        Time.utc(2015, 11, 30, 21, 0, 0) => 1
+      }
     }
     @log_ticket = LogTicket.create!(backend: "s3", directory: "test-bucket", key: "fastly-fake.log", status: "pending")
 
@@ -81,7 +94,7 @@ class FastlyLogProcessorJobTest < ActiveJob::TestCase
           next unless version
           count = GemDownload.find_by(rubygem_id: version.rubygem.id, version_id: version.id).count
 
-          assert_equal expected_count, count, "invalid value for #{name}"
+          assert_equal expected_count.each_value.sum, count, "invalid value for #{name}"
         end
 
         json = Rubygem.find_by_name("json")
@@ -91,6 +104,73 @@ class FastlyLogProcessorJobTest < ActiveJob::TestCase
         assert_equal "processed", @log_ticket.reload.status
       end
 
+      should "update download counts in timescale" do
+        requires_timescale
+
+        @job.perform_now
+
+        downloads = Download.all.map do |download|
+          version = Version.find(download.version_id)
+          { version.full_name => { download.occurred_at.utc => download.downloads } }
+        end.reduce({}, &:deep_merge)
+
+        assert_equal @sample_log_counts.except("no-such-gem-1.2.3"), downloads
+
+        json = Rubygem.find_by_name("json")
+
+        assert_equal 7, es_downloads(json.id)
+        assert_equal "processed", @log_ticket.reload.status
+      end
+
+      context "all the timescale subclasses" do
+        {
+          Download::P1D => [
+            { rubygem: "bundler", version: "1.10.6", downloads: 2, occurred_at: "2015-11-30 00:00:00 UTC" },
+            { rubygem: "json", version: "1.8.3", downloads: 2, occurred_at: "2015-11-30 00:00:00 UTC" },
+            { rubygem: "json", version: "1.8.3", downloads: 1, occurred_at: "2015-11-30 00:00:00 UTC" },
+            { rubygem: "json", version: "1.8.2", downloads: 1, occurred_at: "2015-11-29 00:00:00 UTC" },
+            { rubygem: "json", version: "1.8.2", downloads: 2, occurred_at: "2015-11-30 00:00:00 UTC" },
+            { rubygem: "json", version: "1.8.2", downloads: 1, occurred_at: "2015-12-30 00:00:00 UTC" }
+          ],
+          Download::P1DAllVersion => [
+            { rubygem: "bundler", downloads: 2, occurred_at: "2015-11-30 00:00:00 UTC" },
+            { rubygem: "json", downloads: 1, occurred_at: "2015-11-29 00:00:00 UTC" },
+            { rubygem: "json", downloads: 5, occurred_at: "2015-11-30 00:00:00 UTC" },
+            { rubygem: "json", downloads: 1, occurred_at: "2015-12-30 00:00:00 UTC" }
+          ],
+          Download::P1DAllGem => [
+            { downloads: 1, occurred_at: "2015-11-29 00:00:00 UTC" },
+            { downloads: 7, occurred_at: "2015-11-30 00:00:00 UTC" },
+            { downloads: 1, occurred_at: "2015-12-30 00:00:00 UTC" }
+          ]
+          # Download::P1M => [],
+          # Download::P1MAllVersion => [],
+          # Download::P1MAllGem => [],
+          # Download::P1Y => [],
+          # Download::P1YAllVersion => [],
+          # Download::P1YAllGem => []
+        }.each do |klass, expected|
+          should "return the correct objects for #{klass}" do
+            requires_timescale
+
+            @job.perform_now
+
+            includes = { # rubocop:disable Performance/CollectionLiteralInLoop
+              rubygem: :name,
+              version: :number
+            }.slice(*klass.columns_hash.keys.map { _1.sub(/_id/, "").to_sym })
+            order = %w[rubygem_id version_id occurred_at] & klass.columns_hash.keys # rubocop:disable Performance/CollectionLiteralInLoop
+            actual = klass.order(order).includes(includes.keys)
+            actual = actual.map do |download|
+              includes.to_h { |k, v| [k, download.send(k).send(v)] }
+                .merge(downloads: download.downloads, occurred_at: download.occurred_at.to_s)
+            end
+
+            assert_equal expected, actual
+          end
+        end
+      end
+
       should "not run if already processed" do
         json = Rubygem.find_by_name("json")
 
diff --git a/test/models/download_test.rb b/test/models/download_test.rb
new file mode 100644
index 00000000000..7d8b29c4329
--- /dev/null
+++ b/test/models/download_test.rb
@@ -0,0 +1,7 @@
+require "test_helper"
+
+class DownloadTest < ActiveSupport::TestCase
+  # test "the truth" do
+  #   assert true
+  # end
+end
diff --git a/test/sample_logs/fastly-fake.log b/test/sample_logs/fastly-fake.log
index 6618a5f6210..4dfe8e01203 100644
--- a/test/sample_logs/fastly-fake.log
+++ b/test/sample_logs/fastly-fake.log
@@ -4,8 +4,8 @@
 <134>2015-11-30T21:02:08Z cache-dfw1829 downloads[191209]: 0.0.0.0 Mon, 30 Nov 2015 21:02:01 GMT /production.s3.rubygems.org/gems/json-1.8.3-java.gem 200 Ruby, RubyGems/2.1.9 universal-java-1.8 Ruby/1.9.3 (2015-01-30 patchlevel 551) jruby
 <134>2015-11-30T21:00:19Z cache-atl6227 downloads[191209]: 0.0.0.0 Mon, 30 Nov 2015 21:00:11 GMT /production.s3.rubygems.org/gems/json-1.8.3.gem 200 Ruby, RubyGems/2.2.0 x86_64-linux Ruby/2.1.0 (2013-12-25 patchlevel 0)
 <134>2015-11-30T21:00:37Z cache-ams4136 downloads[191209]: 0.0.0.0 Mon, 30 Nov 2015 21:00:36 GMT /production.s3.rubygems.org/gems/json-1.8.2.gem 200 Ruby, RubyGems/1.8.23 x86_64-linux Ruby/1.9.3 (2012-04-20 patchlevel 194)
-<134>2015-11-30T21:00:51Z cache-nrt6130 downloads[191209]: 0.0.0.0 Mon, 30 Nov 2015 21:00:50 GMT /production.s3.rubygems.org/gems/json-1.8.2.gem 200 Ruby, RubyGems/2.2.2 x86_64-linux Ruby/2.1.7 (2015-08-18 patchlevel 400)
-<134>2015-11-30T21:00:51Z cache-nrt6130 downloads[191209]: 0.0.0.0 Mon, 30 Nov 2015 21:00:50 GMT /production.s3.rubygems.org/gems/json-1.8.2.gem 200 Ruby, RubyGems/2.2.2 x86_64-linux Ruby/2.1.7 (2015-08-18 patchlevel 400)
-<134>2015-11-30T21:00:51Z cache-nrt6130 downloads[191209]: 0.0.0.0 Mon, 30 Nov 2015 21:00:50 GMT /production.s3.rubygems.org/gems/json-1.8.2.gem 304 Ruby, RubyGems/2.2.2 x86_64-linux Ruby/2.1.7 (2015-08-18 patchlevel 400)
+<134>2015-11-29T21:00:51Z cache-nrt6130 downloads[191209]: 0.0.0.0 Mon, 30 Nov 2015 21:00:50 GMT /production.s3.rubygems.org/gems/json-1.8.2.gem 200 Ruby, RubyGems/2.2.2 x86_64-linux Ruby/2.1.7 (2015-08-18 patchlevel 400)
+<134>2015-12-30T21:00:51Z cache-nrt6130 downloads[191209]: 0.0.0.0 Mon, 30 Nov 2015 21:00:50 GMT /production.s3.rubygems.org/gems/json-1.8.2.gem 200 Ruby, RubyGems/2.2.2 x86_64-linux Ruby/2.1.7 (2015-08-18 patchlevel 400)
+<134>2015-11-30T21:39:51Z cache-nrt6130 downloads[191209]: 0.0.0.0 Mon, 30 Nov 2015 21:00:50 GMT /production.s3.rubygems.org/gems/json-1.8.2.gem 304 Ruby, RubyGems/2.2.2 x86_64-linux Ruby/2.1.7 (2015-08-18 patchlevel 400)
 <134>2015-11-30T21:00:51Z cache-nrt6130 downloads[191209]: 0.0.0.0 Mon, 30 Nov 2015 21:00:50 GMT /production.s3.rubygems.org/gems/json-1.8.2.gem 404 Ruby, RubyGems/2.2.2 x86_64-linux Ruby/2.1.7 (2015-08-18 patchlevel 400)
 <134>2015-11-30T21:00:51Z cache-nrt6130 downloads[191209]: 0.0.0.0 Mon, 30 Nov 2015 21:00:50 GMT /production.s3.rubygems.org/gems/no-such-gem-1.2.3.gem 200 Ruby, RubyGems/2.2.2 x86_64-linux Ruby/2.1.7 (2015-08-18 patchlevel 400)
diff --git a/test/test_helper.rb b/test/test_helper.rb
index 055d8833cd4..4b09c03a1b4 100644
--- a/test/test_helper.rb
+++ b/test/test_helper.rb
@@ -101,6 +101,13 @@ def requires_toxiproxy
     skip("Toxiproxy is not running, but was required for this test.")
   end
 
+  def requires_timescale
+    Download.hypertable?
+  rescue StandardError
+    raise "Timescale not running, but REQUIRE_TIMESCALE was set." if ENV["REQUIRE_TIMESCALE"]
+    skip("Timescale is not running, but was required for this test.")
+  end
+
   def assert_changed(object, *attributes)
     original_attributes = attributes.index_with { |a| object.send(a) }
     yield if block_given?