October 2020 - Compass-ci - mailweb.openeuler.org

[PATCH lkp-tests] jobs/build-pkg.yaml: add pkgbuild_repo examples
by Lin Jiaxin 29 Oct '20

29 Oct '20

[why] pkgbuild_repo field's parameter format is not uniform for different git repository. Novices are easy to make mistakes. [how] Add examples of three types of git repository used by build-pkg. Signed-off-by: Lin Jiaxin <ljx.joe(a)qq.com> --- jobs/build-pkg.yaml | 5 +++++ 1 file changed, 5 insertions(+) diff --git a/jobs/build-pkg.yaml b/jobs/build-pkg.yaml index 5888838a..b17fd2ad 100644 --- a/jobs/build-pkg.yaml +++ b/jobs/build-pkg.yaml @@ -1,4 +1,9 @@ suite: build-pkg category: functional +# pkgbuild_repo examples +# pkgbuild_repo: archlinux/aur-j/java8-openjdk +# pkgbuild_repo: archlinux/community//keepalived/trunk +# pkgbuild_repo: archlinux/packages//prometheus-node-exporter/trunk + build-pkg: -- 2.23.0

3 4

[PATCH compass-ci] sched: consume job from queues
by Cao Xueliang 29 Oct '20

29 Oct '20

[why] Before, we consume job from the queue which has relation with tbox_group. Now, we consume job from the queue which is setted by mac2queues. Signed-off-by: Cao Xueliang <caoxl78320(a)163.com> --- src/lib/sched.cr | 69 +++++++++++++++++++++++++++++++++----- src/scheduler/scheduler.cr | 2 +- 2 files changed, 62 insertions(+), 9 deletions(-) diff --git a/src/lib/sched.cr b/src/lib/sched.cr index c32091f..d8d0338 100644 --- a/src/lib/sched.cr +++ b/src/lib/sched.cr @@ -240,10 +240,10 @@ class Sched response = add_job(job, job_id) message = (response["error"]? ? response["error"]["root_cause"] : "") job_messages << { - "job_id" => job_id, - "message" => message.to_s, - "job_state" => "submit", - "result_root" => "/srv#{job.result_root}" + "job_id" => job_id, + "message" => message.to_s, + "job_state" => "submit", + "result_root" => "/srv#{job.result_root}", } return job_messages if response["error"]? end @@ -292,10 +292,10 @@ class Sched message = (response["error"]? ? response["error"]["root_cause"] : "") return [{ - "job_id" => job_id, - "message" => message.to_s, - "job_state" => "submit", - "result_root" => "/srv#{job.result_root}" + "job_id" => job_id, + "message" => message.to_s, + "job_state" => "submit", + "result_root" => "/srv#{job.result_root}", }] end @@ -378,6 +378,59 @@ class Sched end end + def rand_queues(queues) + return queues if queues.empty? + + queues_size = queues.size + base = Random.rand(queues_size) + temp_queues = [] of String + + (0..queues_size - 1).each do |index| + temp_queues << queues[(index + base) % queues_size] + end + + return temp_queues + end + + def get_queues(mac) + queues = [] of String + + queues_str = @redis.hash_get("sched/mac2queues", normalize_mac(mac)) + return queues unless queues_str + + queues_str.split(',', remove_empty: true) do |item| + queues << item.strip + end + + return rand_queues(queues) + end + + def get_job_from_queues(queues, testbox) + job = nil + + queues.each do |queue| + job = prepare_job("sched/#{queue}", testbox) + return job if job + end + + return job + end + + def get_job_boot(env : HTTP::Server::Context) + value = env.params.url["value"] + boot_type = env.params.url["boot_type"] + testbox = @redis.hash_get("sched/mac2host", normalize_mac(value)) + + queues = get_queues(value) + job = get_job_from_queues(queues, testbox) + + if job + Jobfile::Operate.create_job_cpio(job.dump_to_json_any, Kemal.config.public_folder) + end + + return boot_content(job, boot_type) + end + def find_job_boot(env : HTTP::Server::Context) api_param = env.params.url["value"] diff --git a/src/scheduler/scheduler.cr b/src/scheduler/scheduler.cr index 371e4ed..1bfea02 100644 --- a/src/scheduler/scheduler.cr +++ b/src/scheduler/scheduler.cr @@ -51,7 +51,7 @@ module Scheduler # /boot.yyy/mac/${mac} get "/boot.:boot_type/:parameter/:value" do |env| env.response.headers["Connection"] = "close" - response = sched.find_job_boot(env) + response = sched.get_job_boot(env) job_id = response[/tmpfs\/(.*)\/job\.cgz/, 1]? puts %({"job_id": "#{job_id}", "job_state": "boot"}) if job_id -- 2.23.0

3 2

[PATCH compass-ci] delimiter: delete tbox_group settings
by Cao Xueliang 29 Oct '20

29 Oct '20

[why] Use the original tbox_group of the job. Signed-off-by: Cao Xueliang <caoxl78320(a)163.com> --- lib/git_bisect.rb | 4 +--- src/delimiter/constants.rb | 1 - src/delimiter/find-commit/bisect_run_script.rb | 9 +++------ 3 files changed, 4 insertions(+), 10 deletions(-) diff --git a/lib/git_bisect.rb b/lib/git_bisect.rb index d9e5cf4..c548861 100644 --- a/lib/git_bisect.rb +++ b/lib/git_bisect.rb @@ -42,7 +42,6 @@ class GitBisect @bad_job.delete('stats') @bad_job.delete('id') @bad_job.delete('error_ids') - @bad_job['tbox_group'] = DELIMITER_TBOX_GROUP end def set_upstream @@ -75,8 +74,7 @@ class GitBisect result = `git -C #{@work_dir} bisect start #{@upstream_commit} #{@good_commit}` temp = result.split(/\n/) if temp[0].include? 'Bisecting' - result = `git -C #{@work_dir} bisect run #{BISECT_RUN_SCRIPT} #{@bad_job_id} "#{@error_id}" \ - "#{DELIMITER_TBOX_GROUP}" #{@work_dir}` + result = `git -C #{@work_dir} bisect run #{BISECT_RUN_SCRIPT} #{@bad_job_id} "#{@error_id}" #{@work_dir}` end FileUtils.rm_r(@work_dir) if Dir.exist?(@work_dir) puts "\nbisect result: #{result}" diff --git a/src/delimiter/constants.rb b/src/delimiter/constants.rb index dafc2c8..93b633d 100644 --- a/src/delimiter/constants.rb +++ b/src/delimiter/constants.rb @@ -9,4 +9,3 @@ MONITOR_PORT = ENV['MONITOR_PORT'] || '11310' TMEP_GIT_BASE = '/c/public_git' DELIMITER_TASK_QUEUE = 'delimiter' BISECT_RUN_SCRIPT = "#{ENV['CCI_SRC']}/src/delimiter/find-commit/bisect_run_script.rb" -DELIMITER_TBOX_GROUP = 'vm-2p8g--delimiter' diff --git a/src/delimiter/find-commit/bisect_run_script.rb b/src/delimiter/find-commit/bisect_run_script.rb index 539599f..d5e4226 100755 --- a/src/delimiter/find-commit/bisect_run_script.rb +++ b/src/delimiter/find-commit/bisect_run_script.rb @@ -10,18 +10,16 @@ require_relative "#{ENV['CCI_SRC']}/src/delimiter/utils" # git bisect run class GitBisectRun - def initialize(job_id, error_id, tbox_group, work_dir) + def initialize(job_id, error_id, work_dir) @es = ESQuery.new @job_id = job_id @error_id = error_id - @tbox_group = tbox_group @work_dir = work_dir end def git_bisect job = @es.query_by_id @job_id job.delete('stats') if job.key?('stats') - job['tbox_group'] = @tbox_group commit = `git -C #{@work_dir} log --pretty=format:"%H" -1` job['upstream_commit'] = commit get_bisect_status job @@ -41,8 +39,7 @@ end job_id = ARGV[0] error_id = ARGV[1] -tbox_group = ARGV[2] -work_dir = ARGV[3] +work_dir = ARGV[2] -run = GitBisectRun.new job_id, error_id, tbox_group, work_dir +run = GitBisectRun.new job_id, error_id, work_dir run.git_bisect -- 2.23.0

1 0

[PATCH compass-ci] container/rsync-server: modify data upload directory
by Cui Yili 29 Oct '20

29 Oct '20

[why] Modify data upload directory to upload qemu image. [how] Change the rsync.conf configuration and docker operating parameters. Signed-off-by: cuiyili <2268260388(a)qq.com> --- container/rsync-server/rsyncd.conf | 6 ++++-- container/rsync-server/start | 4 ++-- 2 files changed, 6 insertions(+), 4 deletions(-) diff --git a/container/rsync-server/rsyncd.conf b/container/rsync-server/rsyncd.conf index 528e8bd..dd8f128 100644 --- a/container/rsync-server/rsyncd.conf +++ b/container/rsync-server/rsyncd.conf @@ -1,10 +1,12 @@ uid = lkp gid = lkp +port = 11387 use chroot = no +log file=/var/log/rsync.log -[result] -path = /result +[upload] +path = /srv/initrd/qemu-image read only = no max connections = 200 refuse options = delete diff --git a/container/rsync-server/start b/container/rsync-server/start index 4124192..6fa3de2 100755 --- a/container/rsync-server/start +++ b/container/rsync-server/start @@ -9,9 +9,9 @@ cmd=( docker run -id --name=rsync_server - -v /srv/result:/result + -v /srv/initrd/qemu-image:/srv/initrd/qemu-image -v /etc/localtime:/etc/localtime:ro - -p 873:873 + -p 11387:11387 rsync_server:latest ) -- 2.23.0

2 1

[PATCH compass-ci 1/2] auto group jobs with user-defined template
by Lu Weitao 29 Oct '20

29 Oct '20

auto group jobs with user-defined template to support template compare mode background: before compare, create job-groups for all jobs that query from ES with user's filter then, create matrices-groups according job-groups finally, compare each group matrices Signed-off-by: Lu Weitao <luweitaobe(a)163.com> --- lib/params_group.rb | 120 ++++++++++++++++++++++++++++++++++++++++++++ 1 file changed, 120 insertions(+) diff --git a/lib/params_group.rb b/lib/params_group.rb index 0f07d26..4d10dbe 100644 --- a/lib/params_group.rb +++ b/lib/params_group.rb @@ -113,3 +113,123 @@ end def remove_singleton(groups) groups.delete_if { |_k, v| v.length < 2 } end + +# -------------------------------------------------------------------------------------------------- +# auto group job_list by user's template +# Exammple: +# Input: +# 1. jobs_list. +# 2. params from template include: +# groups_params(x_params): +# eg: ['block_size', 'package_size'] +# dimensions: +# eg: [ +# {'os' => 'openeuler', 'os_version' => '20.03'}, +# {'os' => 'centos', 'os_version' => '7.6'} +# ] +# metrics: +# eg: ['fio.read_iops', 'fio_write_iops'] +# Output: +# eg: +# { +# '4k|1G' => { +# 'openeuler 20.03' => [ +# {'stats' => {'fio.write_iops' => 312821.002387, 'fio.read_iops' => 212821.2387}}, +# {'stats' => {'fio.write_iops' => 289661.878453}}, +# ... +# ], +# 'contos 7.6' => [...] +# }, +# '16K|1G' => {...}, +# ... +# } +# + +def auto_group_by_template(jobs_list, group_params, dimensions, metrics) + job_list = extract_jobs_list(jobs_list) + get_group_by_template(job_list, group_params, dimensions, metrics) +end + +def get_group_by_template(job_list, group_params, dimensions, metrics) + groups = {} + job_list.each do |job| + next unless job['stats'] + + group_key = get_user_group_key(job, group_params) + dimension = get_user_dimension(job, dimensions) + next unless group_key && dimension + + new_job = get_new_job_by_metrics(job, metrics) + next if new_job.empty? + + groups[group_key] ||= {} + groups[group_key][dimension] ||= [] + groups[group_key][dimension] << new_job + end + groups +end + +# @group_params Array(String) +# eg: +# ['block_size', 'package_size'] +# return eg: +# '4K|1G' +def get_user_group_key(job, group_params) + group_key_list = [] + group_params.each do |param| + if job.key?(param) + group_key_list << job[param] + next + end + job.each_value do |v| + if v.is_a?(Hash) && v.key?(param) + group_key_list << v[param] + break + end + end + end + return nil if group_key_list.size < group_params.size + + group_key_list.join('|') +end + +# @dimension Array(Hash) +# eg: +# [ +# {os => openeuler, os_version => 20.03}, +# {os => centos, os_version => 7.6} +# ] +# return eg: +# 'openeuler 20.03' +def get_user_dimension(job, dimensions) + dimension_list = [] + + dimensions.each do |dim| + dim.each_key do |key| + if job[key] == dim[key] + dimension_list << job[key] + end + end + break unless dimension_list.empty? + end + return nil if dimension_list.size < dimensions.size + + dimension_list.join(' ') +end + +# @metrics Array(String) +# eg: +# ["fio.read_iops", "fio.write_iops"] +# return new_job +# eg: +# {'stats' => {'fio.write_iops' => 312821.002387, 'fio.read_iops' => 212821.2387}}, +def get_new_job_by_metrics(job, metrics) + new_job = {} + metrics.each do |metric| + if job['stats'].key?(metric) + new_job['stats'] ||= {} + new_job['stats'][metric] = job['stats'][metric] + end + end + new_job +end -- 2.23.0

4 6

[PATCH lab-z9] hosts: delete repet arch field
by Zhang Yu 29 Oct '20

29 Oct '20

Signed-off-by: Zhang Yu <2134782174(a)qq.com> --- hosts/taishan200-2280-2s48p-256g | 1 - hosts/taishan200-2280-2s64p-256g | 1 - 2 files changed, 2 deletions(-) diff --git a/hosts/taishan200-2280-2s48p-256g b/hosts/taishan200-2280-2s48p-256g index 642e8e1..b326025 100644 --- a/hosts/taishan200-2280-2s48p-256g +++ b/hosts/taishan200-2280-2s48p-256g @@ -2,4 +2,3 @@ arch: aarch64 nr_cpu: 96 nr_node: 4 memory: 256G -arch: aarch64 diff --git a/hosts/taishan200-2280-2s64p-256g b/hosts/taishan200-2280-2s64p-256g index 16349d9..7c7eb5c 100644 --- a/hosts/taishan200-2280-2s64p-256g +++ b/hosts/taishan200-2280-2s64p-256g @@ -2,4 +2,3 @@ arch: aarch64 nr_cpu: 128 nr_node: 4 memory: 256G -arch: aarch64 -- 2.23.0

2 3

[PATCH compass-ci 2/2] create matrices-groups according to group jobs
by Lu Weitao 29 Oct '20

29 Oct '20

create matrices-groups according to group jobs for support user-defined template compare mode background: before compare, create job-groups for all jobs that query from ES with user's filter then, create matrices-groups according job-groups finally, compare each group matrices Signed-off-by: Lu Weitao <luweitaobe(a)163.com> --- lib/matrix2.rb | 24 ++++++++++++++++++++++++ 1 file changed, 24 insertions(+) diff --git a/lib/matrix2.rb b/lib/matrix2.rb index 05f96dc..61517d9 100644 --- a/lib/matrix2.rb +++ b/lib/matrix2.rb @@ -120,3 +120,27 @@ def combine_group_query_data(query_data, dims) end groups end + +# input: +# 1. query results from es_query +# 2. params from template include: +# groups_params(x_params): +# eg: ['block_size', 'package_size'] +# dimensions: +# eg: [ +# {'os' => 'openeuler', 'os_version' => '20.03'}, +# {'os' => 'centos', 'os_version' => '7.6'} +# ] +# metrics: +# eg: ['fio.read_iops', 'fio_write_iops'] +# return: group_matrix of Hash(String, Hash(String, matrix)) +def combine_user_group_query_data(query_data, groups_params, dimensions, metrics) + job_list = query_data['hits']['hits'] + groups = auto_group_by_template(job_list, groups_params, dimensions, metrics) + groups.each do |group_key, value| + value.each do |dimension_key, jobs| + groups[group_key][dimension_key] = create_matrix(jobs) + end + end + groups +end -- 2.23.0

3 4

[PATCH compass-ci] container/netdata: use json-file save logs
by Wu Zhende 29 Oct '20

29 Oct '20

[Why] netdata is aperformance check tool and does not need to send logs to fluentd for analysis. Signed-off-by: Wu Zhende <wuzhende666(a)163.com> --- container/netdata/start | 2 ++ 1 file changed, 2 insertions(+) diff --git a/container/netdata/start b/container/netdata/start index a3009a9..58fb675 100755 --- a/container/netdata/start +++ b/container/netdata/start @@ -17,6 +17,8 @@ cmd=( -v /var/run/docker.sock:/var/run/docker.sock:ro -v /srv/cache/netdata_cache:/var/cache/netdata -v /srv/cache/netdata_lib:/var/lib/netdata + --log-driver json-file + --log-opt max-size=1g --cap-add SYS_PTRACE --security-opt apparmor=unconfined netdata/netdatalkp -- 2.23.0

1 0

[PATCH lab-z9] hosts: Delete duplicate field
by Zhang Yu 29 Oct '20

29 Oct '20

Signed-off-by: Zhang Yu <2134782174(a)qq.com> --- hosts/taishan200-2280-2s48p-256g | 1 - hosts/taishan200-2280-2s64p-256g | 1 - 2 files changed, 2 deletions(-) diff --git a/hosts/taishan200-2280-2s48p-256g b/hosts/taishan200-2280-2s48p-256g index 642e8e1..b326025 100644 --- a/hosts/taishan200-2280-2s48p-256g +++ b/hosts/taishan200-2280-2s48p-256g @@ -2,4 +2,3 @@ arch: aarch64 nr_cpu: 96 nr_node: 4 memory: 256G -arch: aarch64 diff --git a/hosts/taishan200-2280-2s64p-256g b/hosts/taishan200-2280-2s64p-256g index 16349d9..7c7eb5c 100644 --- a/hosts/taishan200-2280-2s64p-256g +++ b/hosts/taishan200-2280-2s64p-256g @@ -2,4 +2,3 @@ arch: aarch64 nr_cpu: 128 nr_node: 4 memory: 256G -arch: aarch64 -- 2.23.0

1 0

[PATCH lkp-tests] tests: add get_cpu function
by Zhang Yu 29 Oct '20

29 Oct '20

[why] Obtain more host information from 'lscpu' Signed-off-by: Zhang Yu <2134782174(a)qq.com> --- tests/host-info | 21 ++++++++++++++------- 1 file changed, 14 insertions(+), 7 deletions(-) diff --git a/tests/host-info b/tests/host-info index 46f43f93..84f796dc 100755 --- a/tests/host-info +++ b/tests/host-info @@ -1,18 +1,25 @@ #!/bin/bash -get_node() +get_memory() { - echo "nr_node: $(grep -Eo 'Node [0-9]+,' /proc/zoneinfo | uniq | wc -l)" - echo "nr_cpu: $(nproc)" memory_total=$(cat /proc/meminfo |grep MemTotal | awk '{print $2}') memory_new=$(awk 'BEGIN{printf "%0.2f", '$memory_total'/1024/1024/16}') memory=$(echo `expr $(echo $memory_new | awk '{print int($0)+1}') \* 16`)G echo "memory: $memory" } -get_model_name() +get_cpu() { - echo "model_name: $(lscpu |grep "Model name" | awk '{print $3}')" + echo $1: $(lscpu | awk -F': *' '/'$2':/{print $2}') +} + +get_cpu_info() +{ + cpu_info=("arch Architecture" "nr_node NUMA\snode$s$" "nr_cpu ^CPU$s$" "model_name Model\sname") + for info in "${cpu_info[@]}" + do + get_cpu $info + done } get_nr_disk() @@ -101,11 +108,11 @@ get_ipmi_ip() main() { - get_node + get_memory get_nr_disk get_partitions get_mac - get_model_name + get_cpu_info get_ipmi_ip } -- 2.23.0

1 0