production: Increase loader git's default configuration
This impacts the loaders add-forge-now, add-forge-now-slow and loader-large (which are using the loader git within).
Technically, this opens the new "overrides" entry configuration to configure only the full qualified loader's class name. So only the impacted loader will charge such configuration (even when pods are running with multiple other loaders).
In that case, this specifically configures the loader git (with loader-git's 2.4.0 release) to pass extra parameter of pack files (to accept bigger ones) and the threshold memory size (after which it starts using disk instead).
[1] Diff:
$ /var/tmp/diff-chart.sh adapt-loader-configuration-with-overrides
Switched to branch 'production'
Your branch is up to date with 'origin/production'.
generate config in production branch for values/default.yaml
generate config in production branch for values/production-cassandra.yaml
generate config in production branch for values/production.yaml
generate config in production branch for values/staging-cassandra.yaml
generate config in production branch for values/staging.yaml
Switched to branch 'adapt-loader-configuration-with-overrides'
generate config in adapt-loader-configuration-with-overrides branch for values/default.yaml
generate config in adapt-loader-configuration-with-overrides branch for values/production-cassandra.yaml
generate config in adapt-loader-configuration-with-overrides branch for values/production.yaml
generate config in adapt-loader-configuration-with-overrides branch for values/staging-cassandra.yaml
generate config in adapt-loader-configuration-with-overrides branch for values/staging.yaml
------------- diff for values/default.yaml -------------
No diff
------------- diff for values/production-cassandra.yaml -------------
No diff
------------- diff for values/production.yaml -------------
--- /tmp/production.yaml.before 2023-06-07 13:50:40.472606503 +0200
+++ /tmp/production.yaml.after 2023-06-07 13:50:40.784606735 +0200
@@ -304,40 +304,44 @@
config.yml.template: |
storage:
cls: pipeline
steps:
- cls: buffer
min_batch_size:
content: 1000
content_bytes: 52428800
directory: 1000
directory_entries: 12000
extid: 1000
release: 1000
release_bytes: 52428800
revision: 1000
revision_bytes: 52428800
revision_parents: 2000
- cls: filter
- cls: retry
- cls: remote
url: http://saam.internal.softwareheritage.org:5002
+ overrides:
+ swh.loader.git.loader.GitLoader:
+ pack_size_bytes: 34359738368
+ temp_file_cutoff: 10737418240
celery:
task_broker: amqp://${AMQP_USERNAME}:${AMQP_PASSWORD}@rabbitmq.internal.softwareheritage.org:/
task_acks_late: true
task_queues:
- add_forge_now:swh.loader.git.tasks.UpdateGitRepository
sentry_settings_for_celery_tasks:
__sentry-settings-for-celery-tasks__
metadata_fetcher_credentials:
__metadata-fetcher-credentials__
init-container-entrypoint.sh: |
#!/bin/bash
set -e
CONFIG_FILE=/etc/swh/config.yml
CONFIG_FILE_WIP=/tmp/wip-config.yml
# substitute environment variables when creating the default config.yml
eval echo \""$(</etc/swh/configuration-template/config.yml.template)"\" \
@@ -406,40 +410,44 @@
config.yml.template: |
storage:
cls: pipeline
steps:
- cls: buffer
min_batch_size:
content: 1000
content_bytes: 52428800
directory: 1000
directory_entries: 12000
extid: 1000
release: 1000
release_bytes: 52428800
revision: 1000
revision_bytes: 52428800
revision_parents: 2000
- cls: filter
- cls: retry
- cls: remote
url: http://saam.internal.softwareheritage.org:5002
+ overrides:
+ swh.loader.git.loader.GitLoader:
+ pack_size_bytes: 34359738368
+ temp_file_cutoff: 10737418240
celery:
task_broker: amqp://${AMQP_USERNAME}:${AMQP_PASSWORD}@rabbitmq.internal.softwareheritage.org:/
task_acks_late: true
task_queues:
- add_forge_now_slow:swh.loader.git.tasks.UpdateGitRepository
sentry_settings_for_celery_tasks:
__sentry-settings-for-celery-tasks__
metadata_fetcher_credentials:
__metadata-fetcher-credentials__
init-container-entrypoint.sh: |
#!/bin/bash
set -e
CONFIG_FILE=/etc/swh/config.yml
CONFIG_FILE_WIP=/tmp/wip-config.yml
# substitute environment variables when creating the default config.yml
eval echo \""$(</etc/swh/configuration-template/config.yml.template)"\" \
@@ -610,42 +618,44 @@
config.yml.template: |
storage:
cls: pipeline
steps:
- cls: buffer
min_batch_size:
content: 1000
content_bytes: 52428800
directory: 1000
directory_entries: 12000
extid: 1000
release: 1000
release_bytes: 52428800
revision: 1000
revision_bytes: 52428800
revision_parents: 2000
- cls: filter
- cls: retry
- cls: remote
url: http://saam.internal.softwareheritage.org:5002
- temp_file_cutoff:
- 10737418240
+ overrides:
+ swh.loader.git.loader.GitLoader:
+ pack_size_bytes: 34359738368
+ temp_file_cutoff: 10737418240
celery:
task_broker: amqp://${AMQP_USERNAME}:${AMQP_PASSWORD}@rabbitmq.internal.softwareheritage.org:/
task_acks_late: false
task_queues:
- large_repository:swh.loader.git.tasks.UpdateGitRepository
sentry_settings_for_celery_tasks:
__sentry-settings-for-celery-tasks__
metadata_fetcher_credentials:
__metadata-fetcher-credentials__
init-container-entrypoint.sh: |
#!/bin/bash
set -e
CONFIG_FILE=/etc/swh/config.yml
CONFIG_FILE_WIP=/tmp/wip-config.yml
# substitute environment variables when creating the default config.yml
eval echo \""$(</etc/swh/configuration-template/config.yml.template)"\" \
...snip... (checksum hash change so meh)...
------------- diff for values/staging-cassandra.yaml -------------
No diff
------------- diff for values/staging.yaml -------------
No diffNote: /var/tmp/diff-charts.sh from $1576
@gsamson might be interested ;)
Refs. !62 (closed)
Edited by Antoine R. Dumont