Module: KubernetesMetadata::WatchNamespaces

Includes:
Common
Included in:
Fluent::Plugin::KubernetesMetadataFilter
Defined in:
lib/fluent/plugin/kubernetes_metadata_watch_namespaces.rb

Instance Method Summary collapse

Methods included from Common

#match_annotations, #parse_namespace_metadata, #parse_pod_metadata, #syms_to_strs

Instance Method Details

#get_namespaces_and_start_watcher ⇒ Object

List all namespaces, record the resourceVersion and return a watcher starting from that resourceVersion.



84
85
86
87
88
89
90
91
92
93
94
95
96
97
 
# File 'lib/fluent/plugin/kubernetes_metadata_watch_namespaces.rb', line 84

def get_namespaces_and_start_watcher
  options = {
    resource_version: '0'  # Fetch from API server.
  }
  namespaces = @client.get_namespaces(options)
  namespaces.each do |namespace|
    cache_key = namespace.metadata['uid']
    @namespace_cache[cache_key] = parse_namespace_metadata(namespace)
    @stats.bump(:namespace_cache_host_updates)
  end
  options[:resource_version] = namespaces.resourceVersion
  watcher = @client.watch_namespaces(options)
  watcher
end

#process_namespace_watcher_notices(watcher) ⇒ Object

Process a watcher notice and potentially raise an exception.



107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
 
# File 'lib/fluent/plugin/kubernetes_metadata_watch_namespaces.rb', line 107

def process_namespace_watcher_notices(watcher)
  watcher.each do |notice|
    case notice.type
      when 'MODIFIED'
        reset_namespace_watch_retry_stats
        cache_key = notice.object['metadata']['uid']
        cached    = @namespace_cache[cache_key]
        if cached
          @namespace_cache[cache_key] = parse_namespace_metadata(notice.object)
          @stats.bump(:namespace_cache_watch_updates)
        else
          @stats.bump(:namespace_cache_watch_misses)
        end
      when 'DELETED'
        reset_namespace_watch_retry_stats
        # ignore and let age out for cases where
        # deleted but still processing logs
        @stats.bump(:namespace_cache_watch_deletes_ignored)
      when 'ERROR'
        @stats.bump(:namespace_watch_error_type_notices)
        message = notice['object']['message'] if notice['object'] && notice['object']['message']
        raise "Error while watching namespaces: #{message}"
      else
        reset_namespace_watch_retry_stats
        # Don't pay attention to creations, since the created namespace may not
        # be used by any namespace on this node.
        @stats.bump(:namespace_cache_watch_ignored)
    end
  end
end

#reset_namespace_watch_retry_stats ⇒ Object

Reset namespace watch retry count and backoff interval as there is a successful watch notice.



101
102
103
104
 
# File 'lib/fluent/plugin/kubernetes_metadata_watch_namespaces.rb', line 101

def reset_namespace_watch_retry_stats
  Thread.current[:namespace_watch_retry_count] = 0
  Thread.current[:namespace_watch_retry_backoff_interval] = @watch_retry_interval
end

#set_up_namespace_thread ⇒ Object 



26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
 
# File 'lib/fluent/plugin/kubernetes_metadata_watch_namespaces.rb', line 26

def set_up_namespace_thread
  # Any failures / exceptions in the initial setup should raise
  # Fluent:ConfigError, so that users can inspect potential errors in
  # the configuration.
  namespace_watcher = start_namespace_watch
  Thread.current[:namespace_watch_retry_backoff_interval] = @watch_retry_interval
  Thread.current[:namespace_watch_retry_count] = 0

  # Any failures / exceptions in the followup watcher notice
  # processing will be swallowed and retried. These failures /
  # exceptions could be caused by Kubernetes API being temporarily
  # down. We assume the configuration is correct at this point.
  while true
    begin
      namespace_watcher ||= get_namespaces_and_start_watcher
      process_namespace_watcher_notices(namespace_watcher)
    rescue Exception => e
      @stats.bump(:namespace_watch_failures)
      if Thread.current[:namespace_watch_retry_count] < @watch_retry_max_times
        # Instead of raising exceptions and crashing Fluentd, swallow
        # the exception and reset the watcher.
        log.info(
          "Exception encountered parsing namespace watch event. " \
          "The connection might have been closed. Sleeping for " \
          "#{Thread.current[:namespace_watch_retry_backoff_interval]} " \
          "seconds and resetting the namespace watcher.", e)
        sleep(Thread.current[:namespace_watch_retry_backoff_interval])
        Thread.current[:namespace_watch_retry_count] += 1
        Thread.current[:namespace_watch_retry_backoff_interval] *= @watch_retry_exponential_backoff_base
        namespace_watcher = nil
      else
        # Since retries failed for many times, log as errors instead
        # of info and raise exceptions and trigger Fluentd to restart.
        message =
          "Exception encountered parsing namespace watch event. The " \
          "connection might have been closed. Retried " \
          "#{@watch_retry_max_times} times yet still failing. Restarting."
        log.error(message, e)
        raise Fluent::UnrecoverableError.new(message)
      end
    end
  end
end

#start_namespace_watch ⇒ Object 



70
71
72
73
74
75
76
77
78
79
80
 
# File 'lib/fluent/plugin/kubernetes_metadata_watch_namespaces.rb', line 70

def start_namespace_watch
  return get_namespaces_and_start_watcher
rescue Exception => e
  message = "start_namespace_watch: Exception encountered setting up " \
            "namespace watch from Kubernetes API #{@apiVersion} endpoint " \
            "#{@kubernetes_url}: #{e.message}"
  message += " (#{e.response})" if e.respond_to?(:response)
  log.debug(message)

  raise Fluent::ConfigError, message
end