ArielGlenn has uploaded a new change for review.
https://gerrit.wikimedia.org/r/300861
Change subject: move cron job for pagetitle list generation to snapshot1007
......................................................................
move cron job for pagetitle list generation to snapshot1007
take the opportunity to get rid of params we don't need/use
in these cron job manifests
Bug: T133694
Change-Id: Ibf938978c8e3b1daab293981d5637b23de8aec89
---
M modules/role/manifests/snapshot/cron/primary.pp
M modules/role/manifests/snapshot/cron/secondary.pp
M modules/role/manifests/snapshot/cronjobs.pp
M modules/snapshot/manifests/cron/pagetitles.pp
4 files changed, 7 insertions(+), 27 deletions(-)
git pull ssh://gerrit.wikimedia.org:29418/operations/puppet
refs/changes/61/300861/1
diff --git a/modules/role/manifests/snapshot/cron/primary.pp
b/modules/role/manifests/snapshot/cron/primary.pp
index e4dda5f..dd13ff2 100644
--- a/modules/role/manifests/snapshot/cron/primary.pp
+++ b/modules/role/manifests/snapshot/cron/primary.pp
@@ -13,10 +13,6 @@
enable => true,
user => 'datasets',
}
- class { 'snapshot::cron::pagetitles':
- enable => true,
- user => 'datasets',
- }
class { 'snapshot::addschanges':
enable => true,
user => 'datasets',
diff --git a/modules/role/manifests/snapshot/cron/secondary.pp
b/modules/role/manifests/snapshot/cron/secondary.pp
index 1acad79..4338ba4 100644
--- a/modules/role/manifests/snapshot/cron/secondary.pp
+++ b/modules/role/manifests/snapshot/cron/secondary.pp
@@ -13,10 +13,6 @@
enable => false,
user => 'datasets',
}
- class { 'snapshot::cron::pagetitles':
- enable => false,
- user => 'datasets',
- }
class { 'snapshot::addschanges':
enable => false,
user => 'datasets',
@@ -29,9 +25,6 @@
enable => false,
user => 'datasets',
}
-# class { 'snapshot::cron::mediaperprojectlists':
-# user => 'datasets',
-# }
class { 'snapshot::cron::cirrussearch':
enable => false,
user => 'datasets',
diff --git a/modules/role/manifests/snapshot/cronjobs.pp
b/modules/role/manifests/snapshot/cronjobs.pp
index 7f82a70..19d0aa6 100644
--- a/modules/role/manifests/snapshot/cronjobs.pp
+++ b/modules/role/manifests/snapshot/cronjobs.pp
@@ -1,4 +1,5 @@
class role::snapshot::cronjobs {
include dataset::user
class { '::snapshot::cron::mediaperprojectlists': user => 'datasets' }
+ class { '::snapshot::cron::pagetitles': user => 'datasets' }
}
diff --git a/modules/snapshot/manifests/cron/pagetitles.pp
b/modules/snapshot/manifests/cron/pagetitles.pp
index 9b43d1c..391dde4 100644
--- a/modules/snapshot/manifests/cron/pagetitles.pp
+++ b/modules/snapshot/manifests/cron/pagetitles.pp
@@ -1,18 +1,10 @@
class snapshot::cron::pagetitles(
- $enable=true,
$user=undef,
) {
include snapshot::dumps::dirs
- include snapshot::cron::wikiqueryskip
-
- if ($enable) {
- $ensure = 'present'
- }
- else {
- $ensure = 'absent'
- }
$otherdir = "${snapshot::dumps::dirs::datadir}/public/other"
+ $repodir = $snapshot::dumps::dirs::repodir
file { "${otherdir}/pagetitles":
ensure => 'directory',
@@ -31,7 +23,7 @@
}
cron { 'titles-cleanup':
- ensure => $ensure,
+ ensure => 'present',
environment => '[email protected]',
user => $user,
command => "find ${otherdir}/pagetitles/ -maxdepth 1 -type d
-mtime +90 -exec rm -rf {} \\; ; find ${otherdir}/mediatitles/ -maxdepth 1
-type d -mtime +90 -exec rm -rf {} \\;",
@@ -39,23 +31,21 @@
hour => '8',
}
- $scriptsdir = $snapshot::dumps::dirs::scriptsdir
-
cron { 'pagetitles-ns0':
- ensure => $ensure,
+ ensure => 'present',
environment => '[email protected]',
user => $user,
- command => "cd ${scriptsdir}; python onallwikis.py --configfile
confs/wikidump.conf.monitor --filenameformat '{w}-{d}-all-titles-in-ns-0.gz'
--outdir '${otherdir}/pagetitles/{d}' --query \"'select page_title from page
where page_namespace=0;'\"",
+ command => "cd ${repodir}; python onallwikis.py --configfile
confs/wikidump.conf.monitor --filenameformat '{w}-{d}-all-titles-in-ns-0.gz'
--outdir '${otherdir}/pagetitles/{d}' --query \"'select page_title from page
where page_namespace=0;'\"",
minute => '10',
hour => '8',
require => File["${otherdir}/pagetitles"],
}
cron { 'pagetitles-ns6':
- ensure => $ensure,
+ ensure => 'present',
environment => '[email protected]',
user => $user,
- command => "cd ${scriptsdir}; python onallwikis.py --configfile
confs/wikidump.conf.monitor --filenameformat '{w}-{d}-all-media-titles.gz'
--outdir '${otherdir}/mediatitles/{d}' --query \"'select page_title from page
where page_namespace=6;'\"",
+ command => "cd ${repodir}; python onallwikis.py --configfile
confs/wikidump.conf.monitor --filenameformat '{w}-{d}-all-media-titles.gz'
--outdir '${otherdir}/mediatitles/{d}' --query \"'select page_title from page
where page_namespace=6;'\"",
minute => '50',
hour => '8',
require => File["${otherdir}/mediatitles"],
--
To view, visit https://gerrit.wikimedia.org/r/300861
To unsubscribe, visit https://gerrit.wikimedia.org/r/settings
Gerrit-MessageType: newchange
Gerrit-Change-Id: Ibf938978c8e3b1daab293981d5637b23de8aec89
Gerrit-PatchSet: 1
Gerrit-Project: operations/puppet
Gerrit-Branch: production
Gerrit-Owner: ArielGlenn <[email protected]>
_______________________________________________
MediaWiki-commits mailing list
[email protected]
https://lists.wikimedia.org/mailman/listinfo/mediawiki-commits