Run rasdaemon everywhere, with a prometheus exporter for events

[chef.git] / cookbooks / hardware / recipes / default.rb

diff --git a/cookbooks/hardware/recipes/default.rb b/cookbooks/hardware/recipes/default.rb
index fca8031979c17b7bb7df46f3b69e54abe860c251..a77908cf4c7e4fc439b59ada912a11945749a2fb 100644 (file)
--- a/cookbooks/hardware/recipes/default.rb
+++ b/cookbooks/hardware/recipes/default.rb
@@ -51,7 +51,7 @@ end
 
 units = []
 
-if node[:roles].include?("bytemark") || node[:roles].include?("exonetric")
+if node[:roles].include?("bytemark") || node[:roles].include?("exonetric") || node[:roles].include?("prgmr")
   units << "0"
 end
 
@@ -59,14 +59,29 @@ case manufacturer
 when "HP"
   package "hponcfg"
 
+  execute "update-ilo" do
+    action :nothing
+    command "/usr/sbin/hponcfg -f /etc/ilo-defaults.xml"
+  end
+
+  template "/etc/ilo-defaults.xml" do
+    source "ilo-defaults.xml.erb"
+    owner "root"
+    group "root"
+    mode "644"
+    notifies :run, "execute[update-ilo]"
+  end
+
   package "hp-health" do
     action :install
     notifies :restart, "service[hp-health]"
+    only_if { node[:lsb][:release].to_f < 22.04 }
   end
 
   service "hp-health" do
     action [:enable, :start]
     supports :status => true, :restart => true
+    only_if { node[:lsb][:release].to_f < 22.04 }
   end
 
   if product.end_with?("Gen8", "Gen9")
@@ -141,7 +156,7 @@ if File.exist?("/etc/default/grub")
   execute "update-grub" do
     action :nothing
     command "/usr/sbin/update-grub"
-    not_if { ENV["TEST_KITCHEN"] }
+    not_if { kitchen? }
   end
 
   template "/etc/default/grub" do
@@ -178,8 +193,17 @@ if node[:kernel][:modules].include?("ipmi_si")
   package "ipmitool"
   package "freeipmi-tools"
 
+  template "/etc/prometheus/ipmi_local.yml" do
+    source "ipmi_local.yml.erb"
+    owner "root"
+    group "root"
+    mode "644"
+  end
+
   prometheus_exporter "ipmi" do
     port 9290
+    options "--config.file=/etc/prometheus/ipmi_local.yml"
+    subscribes :restart, "template[/etc/prometheus/ipmi_local.yml]"
   end
 end
 
@@ -190,13 +214,30 @@ service "irqbalance" do
   supports :status => false, :restart => true, :reload => false
 end
 
-# Link Layer Discovery Protocol Daemon
 package "lldpd"
+
 service "lldpd" do
   action [:start, :enable]
   supports :status => true, :restart => true, :reload => true
 end
 
+ohai_plugin "lldp" do
+  template "lldp.rb.erb"
+end
+
+package %w[
+  rasdaemon
+  ruby-sqlite3
+]
+
+service "rasdaemon" do
+  action [:enable, :start]
+end
+
+prometheus_exporter "rasdaemon" do
+  port 9797
+end
+
 tools_packages = []
 status_packages = {}
 
@@ -219,9 +260,6 @@ if node[:virtualization][:role] != "guest" ||
     when "mpt2sas", "mpt3sas"
       tools_packages << "sas2ircu"
       status_packages["sas2ircu-status"] ||= []
-    when "megaraid_mm"
-      tools_packages << "megactl"
-      status_packages["megaraid-status"] ||= []
     when "megaraid_sas"
       tools_packages << "megacli"
       status_packages["megaclisas-status"] ||= []
@@ -265,7 +303,7 @@ if tools_packages.include?("areca")
     depth 1
     user "root"
     group "root"
-    not_if { ENV["TEST_KITCHEN"] }
+    not_if { kitchen? }
   end
 else
   directory "/opt/areca" do
@@ -286,6 +324,7 @@ if status_packages.include?("cciss-vol-status")
   systemd_service "cciss-vol-statusd" do
     description "Check cciss_vol_status values in the background"
     exec_start "/usr/local/bin/cciss-vol-statusd"
+    nice 10
     private_tmp true
     protect_system "full"
     protect_home true
@@ -302,7 +341,7 @@ else
   end
 end
 
-%w[cciss-vol-status mpt-status sas2ircu-status megaraid-status megaclisas-status aacraid-status].each do |status_package|
+%w[cciss-vol-status mpt-status sas2ircu-status megaclisas-status aacraid-status].each do |status_package|
   if status_packages.include?(status_package)
     package status_package
 
@@ -344,28 +383,37 @@ nvmes = if node[:hardware][:pci]
           []
         end
 
+unless nvmes.empty?
+  package "nvme-cli"
+end
+
 intel_nvmes = nvmes.select { |pci| pci[:vendor_name] == "Intel Corporation" }
 
 if !intel_ssds.empty? || !intel_nvmes.empty?
   package "unzip"
 
-  intel_ssd_tool_version = "3.0.26"
+  intel_mas_tool_version = "1.10"
+  intel_mas_package_version = "#{intel_mas_tool_version}.155-0"
 
-  remote_file "#{Chef::Config[:file_cache_path]}/Intel_SSD_Data_Center_Tool_#{intel_ssd_tool_version}_Linux.zip" do
-    source "https://downloadmirror.intel.com/29720/eng/Intel_SSD_DCT_#{intel_ssd_tool_version}_Linux.zip"
+  remote_file "#{Chef::Config[:file_cache_path]}/Intel_MAS_CLI_Tool_#{intel_mas_tool_version}_Linux.zip" do
+    source "https://downloadmirror.intel.com/646992/Intel_MAS_CLI_Tool_Linux_#{intel_mas_tool_version}-v2.zip"
   end
 
-  execute "#{Chef::Config[:file_cache_path]}/Intel_SSD_Data_Center_Tool_#{intel_ssd_tool_version}_Linux.zip" do
-    command "unzip Intel_SSD_Data_Center_Tool_#{intel_ssd_tool_version}_Linux.zip isdct_#{intel_ssd_tool_version}.400-1_amd64.deb"
+  execute "#{Chef::Config[:file_cache_path]}/Intel_MAS_CLI_Tool_#{intel_mas_tool_version}_Linux.zip" do
+    command "unzip Intel_MAS_CLI_Tool_#{intel_mas_tool_version}_Linux.zip intelmas_#{intel_mas_package_version}_amd64.deb"
     cwd Chef::Config[:file_cache_path]
     user "root"
     group "root"
-    not_if { ::File.exist?("#{Chef::Config[:file_cache_path]}/isdct_#{intel_ssd_tool_version}.400-1_amd64.deb") }
+    not_if { ::File.exist?("#{Chef::Config[:file_cache_path]}/intelmas_#{intel_mas_package_version}_amd64.deb") }
+  end
+
+  dpkg_package "intelmas" do
+    version "#{intel_mas_package_version}"
+    source "#{Chef::Config[:file_cache_path]}/intelmas_#{intel_mas_package_version}_amd64.deb"
   end
 
   dpkg_package "isdct" do
-    version "#{intel_ssd_tool_version}-1"
-    source "#{Chef::Config[:file_cache_path]}/isdct_#{intel_ssd_tool_version}.400-1_amd64.deb"
+    action :purge
   end
 end
 
@@ -453,6 +501,18 @@ if disks.count.positive?
     subscribes :restart, "template[/etc/default/smartmontools]"
   end
 
+  template "/etc/prometheus/collectors/smart.devices" do
+    source "smart.devices.erb"
+    owner "root"
+    group "root"
+    mode "644"
+    variables :disks => disks
+  end
+
+  prometheus_collector "smart" do
+    interval "15m"
+  end
+
   # Don't try and do munin monitoring of disks behind
   # an Areca controller as they only allow one thing to
   # talk to the controller at a time and smartd will
@@ -466,18 +526,6 @@ if disks.count.positive?
       conf_variables :disk => disk
     end
   end
-
-  template "/etc/prometheus/collectors/smart.devices" do
-    source "smart.devices.erb"
-    owner "root"
-    group "root"
-    mode "644"
-    variables :disks => disks
-  end
-
-  prometheus_collector "smart" do
-    interval "15m"
-  end
 else
   service "smartd" do
     action [:stop, :disable]
@@ -526,16 +574,21 @@ if File.exist?("/etc/mdadm/mdadm.conf")
   end
 end
 
-template "/etc/modules" do
-  source "modules.erb"
-  owner "root"
-  group "root"
-  mode "644"
+file "/etc/modules" do
+  action :delete
 end
 
-service "kmod" do
-  action :nothing
-  subscribes :start, "template[/etc/modules]"
+node[:hardware][:modules].each do |module_name|
+  kernel_module module_name do
+    action :install
+    not_if { kitchen? }
+  end
+end
+
+node[:hardware][:blacklisted_modules].each do |module_name|
+  kernel_module module_name do
+    action :blacklist
+  end
 end
 
 if node[:hardware][:watchdog]
@@ -613,3 +666,7 @@ if node[:hardware][:shm_size]
     notifies :run, "execute[remount-dev-shm]"
   end
 end
+
+prometheus_collector "ohai" do
+  interval "15m"
+end