ood_core 0.23.4 → 0.24.0

checksums.yaml

@@ -1,7 +1,7 @@
 ---
 SHA256:
-  metadata.gz: aaa540da7d8afffa45674650d11ef7bf260e55ff44e32244950d233e1d4d145a
-  data.tar.gz: ff4e1fbe8c59309f36fc45c634cb27752d71e3e66e847496ab3678f556a90e8b
+  metadata.gz: d5a2f83340f4dad3b9de58e0b5db178e2ec8566da3f28e717405c2692477d49d
+  data.tar.gz: f63183676134dd3b410fa4bf24ea84a3c1e16ccae8efd02bdec0174ec0dee30b
 SHA512:
-  metadata.gz: bc0fecb3a8eae6db9c5d6c40d62d4568f930fc197a153ec28a0942ec2d2bc8da0e488e6fe86d906ce03ca2ae53d911c43804623e7b185335281d23b1cef7af41
-  data.tar.gz: 56eae7c8a31c5f4b625b06b33b84def3d80a0652bd883f39f656a276c63e7224b88cb9ed801ef4482e62d1eeee6aeff366787414d21c52daf03cc4eb2265d83a
+  metadata.gz: 3f356e6fa7d39314ae4092533fcadc7dcd062911875b3d3cf934612015fbdac1be55be66436ffa30ad8e14077c30717fe54ff5d8d4fc7699bc57c9f04ad7570f
+  data.tar.gz: 22c2606e3449c338a3c9dfae4e9f8b5c60a8c172072859dc0aa5ac79be3779dcf337d30a8d351e9273356c74cdc19a235f9267c06aab0405511049971f3fe206

data/CHANGELOG.md

@@ -7,6 +7,27 @@ and this project adheres to [Semantic Versioning](http://semver.org/spec/v2.0.0.
 
 ## [Unreleased]
 
+## [0.24.0] - 11-28-2023
+
+- Code cleanup and separate arguments with whitespace in Fujitsu TCS adapter by @mnakao in https://github.com/OSC/ood_core/pull/808
+- Add OUT_OF_MEMORY state for Slurm by @robinkar in https://github.com/OSC/ood_core/pull/809
+- find_port: avoid infinite loop by @utkarshayachit in https://github.com/OSC/ood_core/pull/811
+- handle find_port error codes by @utkarshayachit in https://github.com/OSC/ood_core/pull/812
+- vnc: run websockify as background process by @utkarshayachit in https://github.com/OSC/ood_core/pull/813
+- Add working_dir option for Fujitsu TCS job scheduler by @mnakao in https://github.com/OSC/ood_core/pull/816
+- Minor fix for Fujitsu TCS by @mnakao in https://github.com/OSC/ood_core/pull/817
+- Update rake requirement from ~> 13.0.1 to ~> 13.1.0 by @dependabot in https://github.com/OSC/ood_core/pull/814
+- Changes default return value for cluster.batch_connect_ssh_allow? by @HazelGrant in https://github.com/OSC/ood_core/pull/818
+
+## [0.23.5] - 04-10-2023
+
+### Fixed
+
+- [804](https://github.com/OSC/ood_core/pull/804) fixed a kubernetes bug in the
+  `info_all` code path.
+- Slurm `-M` flag now correctly accounts for full path `sacctmgr` commands in
+  [807](https://github.com/OSC/ood_core/pull/807).
+
 ## [0.23.4] - 03-06-2023
 
 ### Fixed
@@ -500,8 +521,9 @@ Functionally the same as [0.17.3] but with some CI updates.
 - Initial release!
 
 [Unreleased]: https://github.com/OSC/ood_core/compare/v0.23.4...HEAD
-[0.23.4]: https://github.com/OSC/ood_core/compare/v0.23.4...v0.23.3
-[0.23.3]: https://github.com/OSC/ood_core/compare/v0.23.3...v0.23.2
+[0.23.5]: https://github.com/OSC/ood_core/compare/v0.23.4...v0.23.5
+[0.23.4]: https://github.com/OSC/ood_core/compare/v0.23.3...v0.23.4
+[0.23.3]: https://github.com/OSC/ood_core/compare/v0.23.2...v0.23.3
 [0.23.2]: https://github.com/OSC/ood_core/compare/v0.23.1...v0.23.2
 [0.23.1]: https://github.com/OSC/ood_core/compare/v0.23.0...v0.23.1
 [0.23.0]: https://github.com/OSC/ood_core/compare/v0.22.0...v0.23.0

data/lib/ood_core/batch_connect/template.rb

@@ -162,14 +162,28 @@ module OodCore
                 export -f port_used
 
                 # Find available port in range [$2..$3] for host $1
-                # Default: [#{min_port}..#{max_port}]
+                # Default host: localhost
+                # Default port range: [#{min_port}..#{max_port}]
+                # returns error code (0: success, 1: failed)
+                # On success, the chosen port is echoed on stdout.
                 find_port () {
                   local host="${1:-localhost}"
-                  local port=$(random_number "${2:-#{min_port}}" "${3:-#{max_port}}")
-                  while port_used "${host}:${port}"; do
-                    port=$(random_number "${2:-#{min_port}}" "${3:-#{max_port}}")
+                  local min_port=${2:-#{min_port}}
+                  local max_port=${3:-#{max_port}}
+                  local port_range=($(shuf -i ${min_port}-${max_port}))
+                  local retries=1 # number of retries over the port range if first attempt fails
+                  for ((attempt=0; attempt<=$retries; attempt++)); do
+                    for port in "${port_range[@]}"; do
+                      if port_used "${host}:${port}"; then
+                        continue
+                      fi
+                      echo "${port}"
+                      return 0 # success
+                    done
                   done
-                  echo "${port}"
+
+                  echo "error: failed to find available port in range ${min_port}..${max_port}" >&2
+                  return 1 # failure
                 }
                 export -f find_port
 

data/lib/ood_core/batch_connect/templates/vnc.rb

@@ -134,10 +134,45 @@ module OodCore
             <<-EOT.gsub(/^ {14}/, "")
               #{super}
 
+              # launches websockify in the background; waiting until the process
+              # has started proxying successfully.
+              start_websockify() {
+                local log_file="./websockify.log"
+                # launch websockify in background and redirect all output to a file.
+                #{websockify_cmd} $1 $2 &> $log_file &
+                local ws_pid=$!
+                local counter=0
+
+                # wait till websockify has successfully started
+                echo "[websockify]: pid: $ws_pid (proxying $1 ==> $2)" >&2
+                echo "[websockify]: log file: $log_file" >&2
+                echo "[websockify]: waiting ..." >&2
+                until grep -q -i "proxying from :$1" $log_file
+                do
+                  if ! ps $ws_pid > /dev/null; then
+                    echo "[websockify]: failed to launch!" >&2
+                    return 1
+                  elif [ $counter -ge 5 ]; then
+                    # timeout after ~5 seconds
+                    echo "[websockify]: timed-out :(!" >&2
+                    return 1
+                  else
+                    sleep 1
+                    ((counter=counter+1))
+                  fi
+                done
+                echo "[websockify]: started successfully (proxying $1 ==> $2)" >&2
+                echo $ws_pid
+                return 0
+              }
+
               # Launch websockify websocket server
               echo "Starting websocket server..."
               websocket=$(find_port)
-              #{websockify_cmd} -D ${websocket} localhost:${port}
+              [ $? -eq 0 ] || clean_up 1 # give up if port not found
+
+              ws_pid=$(start_websockify ${websocket} localhost:${port})
+              [ $? -eq 0 ] || clean_up 1 # give up if websockify launch failed
 
               # Set up background process that scans the log file for successful
               # connections by users, and change the password after every

data/lib/ood_core/batch_connect/templates/vnc_container.rb

@@ -173,6 +173,7 @@ module OodCore
               module load #{container_module}
               echo "Starting websocket server..."
               websocket=$(find_port)
+              [ $? -eq 0 ] || clean_up 1 # give up if port not found
               #{container_command} exec instance://#{@instance_name} #{websockify_cmd} -D ${websocket} localhost:${port}
 
               # Set up background process that scans the log file for successful

data/lib/ood_core/cluster.rb

@@ -165,9 +165,9 @@ module OodCore
     # @return [Boolean, nil] whether cluster supports SSH to batch connect node
     def batch_connect_ssh_allow?
       return @batch_connect_ssh_allow if defined?(@batch_connect_ssh_allow)
-      return @batch_connect_ssh_allow = nil if batch_connect_config.nil?
+      return @batch_connect_ssh_allow = true if batch_connect_config.nil?
 
-      @batch_connect_ssh_allow = batch_connect_config.fetch(:ssh_allow, nil)
+      @batch_connect_ssh_allow = batch_connect_config.fetch(:ssh_allow, true)
     end
 
     # The comparison operator

data/lib/ood_core/job/adapters/fujitsu_tcs.rb

@@ -12,11 +12,13 @@ module OodCore
       # @param config [#to_h] the configuration for job adapter
       # @option config [Object] :bin (nil) Path to Fujitsu TCS resource manager binaries
       # @option config [#to_h]  :bin_overrides ({}) Optional overrides to Fujitsu TCS resource manager executables
+      # @option config [Object] :working_dir (nil) Working directory for submitting a batch script
       def self.build_fujitsu_tcs(config)
         c = config.to_h.symbolize_keys
         bin           = c.fetch(:bin, nil)
         bin_overrides = c.fetch(:bin_overrides, {})
-        fujitsu_tcs   = Adapters::Fujitsu_TCS::Batch.new(bin: bin, bin_overrides: bin_overrides)
+        working_dir   = c.fetch(:working_dir, nil)
+        fujitsu_tcs   = Adapters::Fujitsu_TCS::Batch.new(bin: bin, bin_overrides: bin_overrides, working_dir: working_dir)
         Adapters::Fujitsu_TCS.new(fujitsu_tcs: fujitsu_tcs)
       end
     end
@@ -43,6 +45,11 @@ module OodCore
           # @return Hash<String, String>
           attr_reader :bin_overrides
 
+          # Working directory for submitting a batch script
+          # @example
+          #   my_batch.working_dir #=> "HOME" or Dir.pwd
+          attr_reader :working_dir
+
           # The root exception class that all Fujitsu TCS specific exceptions inherit
           # from
           class Error < StandardError; end
@@ -52,9 +59,17 @@ module OodCore
 
           # @param bin [#to_s] path to Fujitsu TCS installation binaries
           # @param bin_overrides [#to_h] a hash of bin ovverides to be used in job
-          def initialize(bin: nil, bin_overrides: {})
+          # @param working_dir [] Working directory for submitting a batch script
+          def initialize(bin: nil, bin_overrides: {}, working_dir: nil)
             @bin           = Pathname.new(bin.to_s)
             @bin_overrides = bin_overrides
+            if working_dir == nil
+              @working_dir = Dir.pwd
+            elsif working_dir == "HOME"
+              @working_dir   = Dir.home
+            else
+              raise(StandardError, "Unknown working_dir")
+            end
           end
 
           # Get a list of hashes detailing each of the jobs on the batch server
@@ -79,19 +94,19 @@ module OodCore
           # @raise [Error] if `pjstat` command exited unsuccessfully
           # @return [Array<Hash>] list of details for jobs
           def get_jobs(id: "", owner: nil)
-            args = ["-s", "--data", "--choose=jid,jnam,rscg,st,std,stde,adt,sdt,nnumr,usr,elpl,elp"]
-            args.concat ["--filter jid=" + id.to_s] unless id.to_s.empty?
-            args.concat ["--filter usr=" + owner.to_s] unless owner.to_s.empty?
+            args = ["-A", "-s", "--data", "--choose=jid,jnam,rscg,st,std,stde,adt,sdt,nnumr,usr,elpl,elp"]
+            args.concat ["--filter", "jid=" + id.to_s] unless id.to_s.empty?
+            args.concat ["--filter", "usr=" + owner.to_s] unless owner.to_s.empty?
             
             StringIO.open(call("pjstat", *args)) do |output|
               output.gets() # Skip header
               jobs = []
               output.each_line do |line|
                 l = line.split(",")
-                jobs << {:JOB_ID => l[1],  :JOB_NAME   => l[2],  :RSC_GRP    => l[3].split(" ")[0],
+                jobs << {:JOB_ID => l[1],  :JOB_NAME   => l[2],  :RSC_GRP    => l[3].split[0],
                          :ST     => l[4],  :STD        => l[5],  :STDE       => l[6],
                          :ACCEPT => l[7],  :START_DATE => l[8],  :NODES      => l[9].split(":")[0],
-                         :USER   => l[10], :ELAPSE_LIM => l[11], :ELAPSE_TIM => l[12].split(" ")[0] }
+                         :USER   => l[10], :ELAPSE_LIM => l[11], :ELAPSE_TIM => l[12].split[0] }
               end
               jobs
             end
@@ -136,16 +151,18 @@ module OodCore
           # @return [String] the id of the job that was created
           def submit_string(str, args: [])
             args = args.map(&:to_s)
-            call("pjsub", *args, stdin: str.to_s).split(" ")[5]
+            call("pjsub", *args, stdin: str.to_s).split[5]
           end
           
           private
             # Call a forked Fujitsu TCS command
             def call(cmd, *args, stdin: "")
               cmd = OodCore::Job::Adapters::Helper.bin_path(cmd, bin, bin_overrides)
-              args  = args.map(&:to_s)
-              o, e, s = Open3.capture3(cmd, *(args.map(&:to_s)), stdin_data: stdin.to_s)
-              s.success? ? o : raise(Error, e)
+              args = args.map(&:to_s)
+              Dir.chdir(working_dir) do
+                o, e, s = Open3.capture3(cmd, *(args.map(&:to_s)), stdin_data: stdin.to_s)
+                s.success? ? o : raise(Error, e)
+              end
             end
         end
 
@@ -221,12 +238,12 @@ module OodCore
           else
             args.concat ["-e", script.error_path]
           end
-          args.concat ["-L rscgrp=" + script.queue_name] unless script.queue_name.nil?
+          args.concat ["-L", "rscgrp=" + script.queue_name] unless script.queue_name.nil?
           args.concat ["-p", script.priority] unless script.priority.nil?
           
           # start_time: <%= Time.local(2023,11,22,13,4).to_i %> in form.yml.erb
           args.concat ["--at", script.start_time.localtime.strftime("%C%y%m%d%H%M")] unless script.start_time.nil?
-          args.concat ["-L elapse=" + seconds_to_duration(script.wall_time)] unless script.wall_time.nil?
+          args.concat ["-L", "elapse=" + seconds_to_duration(script.wall_time)] unless script.wall_time.nil?
           args.concat ["--bulk", "--sparam", script.job_array_request] unless script.job_array_request.nil?
 
           # Set environment variables
@@ -368,7 +385,7 @@ module OodCore
         private
           # Convert duration to seconds
           def duration_in_seconds(time)
-            return 0 if time.nil?
+            return 0 if time.nil? or time == "-"
             time, days = time.split("-").reverse
             days.to_i * 24 * 3600 +
               time.split(':').map { |v| v.to_i }.inject(0) { |total, v| total * 60 + v }

data/lib/ood_core/job/adapters/kubernetes/batch.rb

@@ -306,10 +306,7 @@ class OodCore::Job::Adapters::Kubernetes::Batch
 
   def pod_info_from_json(pod)
     hash = helper.pod_info_from_json(pod)
-    K8sJobInfo.new(hash)
-  rescue Helper::K8sDataError
-    # FIXME: silently eating error, could probably use a logger
-    nil
+    OodCore::Job::Adapters::Kubernetes::K8sJobInfo.new(hash)
   end
 
   def configure_auth(auth)

data/lib/ood_core/job/adapters/slurm.rb

@@ -377,7 +377,7 @@ module OodCore
               cmd = OodCore::Job::Adapters::Helper.bin_path(cmd, bin, bin_overrides)
 
               args  = args.map(&:to_s)
-              args.concat ["-M", cluster] if cluster && cmd != 'sacctmgr'
+              args.concat ["-M", cluster] if cluster && !cmd.to_s.end_with?('sacctmgr')
 
               env = env.to_h
               env["SLURM_CONF"] = conf.to_s if conf
@@ -436,7 +436,8 @@ module OodCore
           'SE' => :completed,  # SPECIAL_EXIT
           'ST' => :running,    # STOPPED
           'S'  => :suspended,  # SUSPENDED
-          'TO' => :completed   # TIMEOUT
+          'TO' => :completed,   # TIMEOUT
+          'OOM' => :completed   # OUT_OF_MEMORY
         }
 
         # @api private

data/lib/ood_core/version.rb

@@ -1,4 +1,4 @@
 module OodCore
   # The current version of {OodCore}
-  VERSION = "0.23.4"
+  VERSION = "0.24.0"
 end

data/ood_core.gemspec

@@ -26,7 +26,7 @@ Gem::Specification.new do |spec|
   spec.add_runtime_dependency "ffi", "~> 1.9", ">= 1.9.6"
   spec.add_runtime_dependency "rexml", "~> 3.2"
   spec.add_development_dependency "bundler", "~> 2.1"
-  spec.add_development_dependency "rake", "~> 13.0.1"
+  spec.add_development_dependency "rake", "~> 13.1.0"
   spec.add_development_dependency "rspec", "~> 3.0"
   spec.add_development_dependency "pry", "~> 0.10"
   spec.add_development_dependency "timecop", "~> 0.8"

metadata

@@ -1,7 +1,7 @@
 --- !ruby/object:Gem::Specification
 name: ood_core
 version: !ruby/object:Gem::Version
-  version: 0.23.4
+  version: 0.24.0
 platform: ruby
 authors:
 - Eric Franz
@@ -10,7 +10,7 @@ authors:
 autorequire: 
 bindir: exe
 cert_chain: []
-date: 2023-03-06 00:00:00.000000000 Z
+date: 2023-11-28 00:00:00.000000000 Z
 dependencies:
 - !ruby/object:Gem::Dependency
   name: ood_support
@@ -80,14 +80,14 @@ dependencies:
     requirements:
     - - "~>"
       - !ruby/object:Gem::Version
-        version: 13.0.1
+        version: 13.1.0
   type: :development
   prerelease: false
   version_requirements: !ruby/object:Gem::Requirement
     requirements:
     - - "~>"
       - !ruby/object:Gem::Version
-        version: 13.0.1
+        version: 13.1.0
 - !ruby/object:Gem::Dependency
   name: rspec
   requirement: !ruby/object:Gem::Requirement