Ottomata has submitted this change and it was merged.
Change subject: Mail webrequest partition status summaries to analytics ops
......................................................................
Mail webrequest partition status summaries to analytics ops
Change-Id: Ieb575dceb491bb0e36f8dd113830f9e98faefe42
---
M manifests/role/analytics/refinery.pp
M manifests/site.pp
2 files changed, 34 insertions(+), 3 deletions(-)
Approvals:
Ottomata: Looks good to me, approved
jenkins-bot: Verified
diff --git a/manifests/role/analytics/refinery.pp
b/manifests/role/analytics/refinery.pp
index c7f8f8e..42f502e 100644
--- a/manifests/role/analytics/refinery.pp
+++ b/manifests/role/analytics/refinery.pp
@@ -82,7 +82,7 @@
}
}
-# == Class role::analytics::refinery::data::check
+# == Class role::analytics::refinery::data::check::icinga
# Configures passive/freshness icinga checks or data imports
# in HDFS.
#
@@ -97,7 +97,7 @@
# See: https://phabricator.wikimedia.org/T76414
# https://bugs.debian.org/cgi-bin/bugreport.cgi?bug=670373
#
-class role::analytics::refinery::data::check {
+class role::analytics::refinery::data::check::icinga {
# We are monitoring hourly datasets.
# Give Oozie a little time to finish running
# the monitor_done_flag workflow for each hour.
@@ -144,3 +144,30 @@
retries => 1,
}
}
+
+# == Class role::analytics::refinery::data::check::email
+# Configures cron jobs that send email about the faultyness of webrequest data
+#
+# These checks walk HDFS through the plain file system.
+#
+class role::analytics::refinery::data::check::email {
+ require role::analytics::refinery
+
+ # This should not be hardcoded. Instead, one should be able to use
+ # $::cdh::hadoop::mount::mount_point to reference the user supplied
+ # parameter when the cdh::hadoop::mount class is evaluated.
+ # I am not sure why this is not working.
+ $hdfs_mount_point = '/mnt/hdfs'
+
+ # Since the 'stats' user is not in ldap, it is unnecessarily hard
+ # to grant it access to the private data in hdfs. As discussed in
+ # https://gerrit.wikimedia.org/r/#/c/186254
+ # the cron runs as hdfs instead.
+ cron { 'refinery data check hdfs_mount':
+ command =>
"${::role::analytics::refinery::path}/bin/refinery-dump-status-webrequest-partitions
--hdfs-mount ${hdfs_mount_point} --quiet",
+ environment => '[email protected],[email protected]',
+ user => 'hdfs',
+ hour => 10,
+ minute => 0,
+ }
+}
diff --git a/manifests/site.pp b/manifests/site.pp
index 109341a..92a0271 100644
--- a/manifests/site.pp
+++ b/manifests/site.pp
@@ -299,7 +299,7 @@
# These are passive checks, so if
# icinga is not notified of a successful import
# hourly, icinga should generate an alert.
- include role::analytics::refinery::data::check
+ include role::analytics::refinery::data::check::icinga
}
@@ -2267,6 +2267,10 @@
# to public data generated by the Analytics Cluster.
include role::analytics::rsyncd
+ # Include analytics/refinery checks that send email about
+ # webrequest partitions faultyness.
+ include role::analytics::refinery::data::check::email
+
# Include the MySQL research password at
# /etc/mysql/conf.d/analytics-research-client.cnf
# and only readable by users in the
--
To view, visit https://gerrit.wikimedia.org/r/185442
To unsubscribe, visit https://gerrit.wikimedia.org/r/settings
Gerrit-MessageType: merged
Gerrit-Change-Id: Ieb575dceb491bb0e36f8dd113830f9e98faefe42
Gerrit-PatchSet: 8
Gerrit-Project: operations/puppet
Gerrit-Branch: production
Gerrit-Owner: QChris <[email protected]>
Gerrit-Reviewer: Gage <[email protected]>
Gerrit-Reviewer: Ottomata <[email protected]>
Gerrit-Reviewer: QChris <[email protected]>
Gerrit-Reviewer: jenkins-bot <>
_______________________________________________
MediaWiki-commits mailing list
[email protected]
https://lists.wikimedia.org/mailman/listinfo/mediawiki-commits
