Ottomata has uploaded a new change for review.
https://gerrit.wikimedia.org/r/75342
Change subject: Moving as many easy filters to oxygen. gadolinium's socat
process is dropping packets with udp2log running.
......................................................................
Moving as many easy filters to oxygen. gadolinium's socat process is dropping
packets with udp2log running.
Change-Id: Ib0c7b8bf3925b828739b80726efe5210c57671a2
---
M manifests/misc/statistics.pp
M templates/udp2log/filters.gadolinium.erb
M templates/udp2log/filters.oxygen.erb
3 files changed, 79 insertions(+), 46 deletions(-)
git pull ssh://gerrit.wikimedia.org:29418/operations/puppet
refs/changes/42/75342/1
diff --git a/manifests/misc/statistics.pp b/manifests/misc/statistics.pp
index cd86847..fe7a426 100644
--- a/manifests/misc/statistics.pp
+++ b/manifests/misc/statistics.pp
@@ -690,21 +690,21 @@
destination => "/a/squid/archive/arabic-banner",
}
- # sampled-1000 logs from gadolinium
+ # sampled-1000 logs from emery
misc::statistics::rsync_job { "sampled_1000":
- source =>
"gadolinium.wikimedia.org::udp2log/webrequest/archive/sampled-1000*.gz",
+ source =>
"emery.wikimedia.org::udp2log/webrequest/archive/sampled-1000*.gz",
destination => "/a/squid/archive/sampled",
}
- # edit logs from gadolinium
+ # edit logs from oxygen
misc::statistics::rsync_job { "edits":
- source =>
"gadolinium.wikimedia.org::udp2log/webrequest/archive/edits*.gz",
+ source =>
"oxygen.wikimedia.org::udp2log/webrequest/archive/edits*.gz",
destination => "/a/squid/archive/edits",
}
- # mobile logs from gadolinium
+ # mobile logs from oxygen
misc::statistics::rsync_job { "mobile":
- source =>
"gadolinium.wikimedia.org::udp2log/webrequest/archive/mobile*.gz",
+ source =>
"oxygen.wikimedia.org::udp2log/webrequest/archive/mobile*.gz",
destination => "/a/squid/archive/mobile",
}
diff --git a/templates/udp2log/filters.gadolinium.erb
b/templates/udp2log/filters.gadolinium.erb
index 9b4eb9b..e954fb5 100644
--- a/templates/udp2log/filters.gadolinium.erb
+++ b/templates/udp2log/filters.gadolinium.erb
@@ -5,37 +5,6 @@
### udp2log packet loss monitoring
pipe 10 /usr/bin/packet-loss 10 '\t' >> <%= webrequest_log_directory
%>/packet-loss.log
-### 0.0001 of all udp2log messages
-# This log file is also on emery for redundancy
-file 1000 <%= webrequest_log_directory %>/sampled-1000.tsv.log
-
-### All edits
-pipe 1 /usr/bin/udp-filter -F '\t' -p action=submit,action=edit >> <%=
webrequest_log_directory %>/edits.tsv.log
-
-
-<%
-# pull in $role::cache::configuration::active_nodes
-# to find mobile host names and build a regex on which to grep.
-cache_configuration =
scope.lookupvar('::role::cache::configuration::active_nodes')
-mobile_hosts_regex = '(' +
cache_configuration['production']['mobile'].values.flatten.join('|') + ')'
--%>
-
-### Mobile traffic filter
-# All mobile traffic goes through CC-cp1044 varnish hosts.
-pipe 100 /bin/grep -P '<%= mobile_hosts_regex %>' >> <%=
webrequest_log_directory %>/mobile-sampled-100.tsv.log
-
-### Fundraising
-## Landing pages
-pipe 1 /usr/bin/udp-filter -F '\t' -d
wikimediafoundation.org,donate.wikimedia.org >> <%= fundraising_log_directory
%>/logs/landingpages.tsv.log
-
-## Banner Impressions
-pipe 100 /usr/bin/udp-filter -F '\t' -p
Special:RecordImpression\?banner=,Special:RecordImpression\?result= >> <%=
fundraising_log_directory %>/logs/bannerImpressions-sampled100.tsv.log
-
-
-
-### All 5xx error responses -- domas (now using udp-filter instead of
5xx-filter).
-# pipe 1 <%= webrequest_filter_directory %>/5xx-filter | awk -W interactive
'$9 !~ "upload.wikimedia.org|query.php"' >> <%= webrequest_log_directory
%>/5xx.tsv.log
-pipe 1 /usr/bin/udp-filter -F '\t' -r -s '^5' | awk -W interactive '$9 !~
"upload.wikimedia.org|query.php"' >> <%= webrequest_log_directory %>/5xx.tsv.log
### webstatscollector -- domas
# NOTE! The files that filter and collector generates
@@ -45,14 +14,48 @@
pipe 1 /usr/local/bin/filter | log2udp -h 127.0.0.1 -p 3815
+### Fundraising
+## Landing pages
+pipe 1 /usr/bin/udp-filter -F '\t' -d
wikimediafoundation.org,donate.wikimedia.org >> <%= fundraising_log_directory
%>/logs/landingpages.tsv.log
-### Vrije Universiteit
-# Contact: <%=
scope.lookupvar('contacts::udp2log::vrije_universiteit_contact') %>
-pipe 10 awk -f <%= webrequest_filter_directory %>/vu.awk | log2udp -h
130.37.198.252 -p 9999
+## Banner Impressions
+pipe 100 /usr/bin/udp-filter -F '\t' -p
Special:RecordImpression\?banner=,Special:RecordImpression\?result= >> <%=
fundraising_log_directory %>/logs/bannerImpressions-sampled100.tsv.log
-### University of Minnesota
-# Contact: <%=
scope.lookupvar('contacts::udp2log::university_minnesota_contact') %>
-# Former Contact: <%=
scope.lookupvar('contacts::udp2log::university_minnesota_contact_former') %>
-# Former contact: <%=
scope.lookupvar('contacts::udp2log::university_minnesota_contact_former2') %>
-pipe 10 awk -f <%= webrequest_filter_directory %>/minnesota.awk | log2udp -h
bento.cs.umn.edu -p 9999
+
+# ----- The filters below have been moved to oxygen -----
+
+<%
+# pull in $role::cache::configuration::active_nodes
+# to find mobile host names and build a regex on which to grep.
+cache_configuration =
scope.lookupvar('::role::cache::configuration::active_nodes')
+mobile_hosts_regex = '(' +
cache_configuration['production']['mobile'].values.flatten.join('|') + ')'
+-%>
+#
+# ### Mobile traffic filter
+# # All mobile traffic goes through CC-cp1044 varnish hosts.
+# pipe 100 /bin/grep -P '<%= mobile_hosts_regex %>' >> <%=
webrequest_log_directory %>/mobile-sampled-100.tsv.log
+#
+#
+# ### 0.0001 of all udp2log messages
+# # This log file is also on emery for redundancy
+# file 1000 <%= webrequest_log_directory %>/sampled-1000.tsv.log
+#
+#
+# ### All 5xx error responses -- domas (now using udp-filter instead of
5xx-filter).
+# # pipe 1 <%= webrequest_filter_directory %>/5xx-filter | awk -W interactive
'$9 !~ "upload.wikimedia.org|query.php"' >> <%= webrequest_log_directory
%>/5xx.tsv.log
+# pipe 1 /usr/bin/udp-filter -F '\t' -r -s '^5' | awk -W interactive '$9 !~
"upload.wikimedia.org|query.php"' >> <%= webrequest_log_directory %>/5xx.tsv.log
+#
+#
+# ### All edits
+# pipe 1 /usr/bin/udp-filter -F '\t' -p action=submit,action=edit >> <%=
webrequest_log_directory %>/edits.tsv.log
+#
+# ### Vrije Universiteit
+# # Contact: <%=
scope.lookupvar('contacts::udp2log::vrije_universiteit_contact') %>
+# pipe 10 awk -f <%= webrequest_filter_directory %>/vu.awk | log2udp -h
130.37.198.252 -p 9999
+#
+# ### University of Minnesota
+# # Contact: <%=
scope.lookupvar('contacts::udp2log::university_minnesota_contact') %>
+# # Former Contact: <%=
scope.lookupvar('contacts::udp2log::university_minnesota_contact_former') %>
+# # Former contact: <%=
scope.lookupvar('contacts::udp2log::university_minnesota_contact_former2') %>
+# pipe 10 awk -f <%= webrequest_filter_directory %>/minnesota.awk | log2udp -h
bento.cs.umn.edu -p 9999
diff --git a/templates/udp2log/filters.oxygen.erb
b/templates/udp2log/filters.oxygen.erb
index 0f29281..1e48033 100644
--- a/templates/udp2log/filters.oxygen.erb
+++ b/templates/udp2log/filters.oxygen.erb
@@ -7,4 +7,34 @@
# Capture all logs with 'zero=' set. The X-Analytics header is set with this
# by mobile varnish frontends upon getting a Wikipedia Zero request.
-pipe 1 /bin/grep -P 'zero=\d{3}-\d{2}' >> /a/log/webrequest/zero.tsv.log
+pipe 1 /bin/grep -P 'zero=\d{3}-\d{2}' >> <%= log_directory %>/zero.tsv.log
+
+### All edits
+pipe 1 /usr/bin/udp-filter -F '\t' -p action=submit,action=edit >> <%=
log_directory %>/edits.tsv.log
+
+<%
+# pull in $role::cache::configuration::active_nodes
+# to find mobile host names and build a regex on which to grep.
+cache_configuration =
scope.lookupvar('::role::cache::configuration::active_nodes')
+mobile_hosts_regex = '(' +
cache_configuration['production']['mobile'].values.flatten.join('|') + ')'
+-%>
+
+### Mobile traffic filter
+# All mobile traffic goes through CC-cp1044 varnish hosts.
+pipe 100 /bin/grep -P '<%= mobile_hosts_regex %>' >> <%= log_directory
%>/mobile-sampled-100.tsv.log
+
+### All 5xx error responses -- domas (now using udp-filter instead of
5xx-filter).
+# pipe 1 <%= webrequest_filter_directory %>/5xx-filter | awk -W interactive
'$9 !~ "upload.wikimedia.org|query.php"' >> <%= log_directory %>/5xx.tsv.log
+pipe 1 /usr/bin/udp-filter -F '\t' -r -s '^5' | awk -W interactive '$9 !~
"upload.wikimedia.org|query.php"' >> <%= log_directory %>/5xx.tsv.log
+
+
+### Vrije Universiteit
+# Contact: <%=
scope.lookupvar('contacts::udp2log::vrije_universiteit_contact') %>
+pipe 10 awk -f <%= webrequest_filter_directory %>/vu.awk | log2udp -h
130.37.198.252 -p 9999
+
+### University of Minnesota
+# Contact: <%=
scope.lookupvar('contacts::udp2log::university_minnesota_contact') %>
+# Former Contact: <%=
scope.lookupvar('contacts::udp2log::university_minnesota_contact_former') %>
+# Former contact: <%=
scope.lookupvar('contacts::udp2log::university_minnesota_contact_former2') %>
+pipe 10 awk -f <%= webrequest_filter_directory %>/minnesota.awk | log2udp -h
bento.cs.umn.edu -p 9999
+
--
To view, visit https://gerrit.wikimedia.org/r/75342
To unsubscribe, visit https://gerrit.wikimedia.org/r/settings
Gerrit-MessageType: newchange
Gerrit-Change-Id: Ib0c7b8bf3925b828739b80726efe5210c57671a2
Gerrit-PatchSet: 1
Gerrit-Project: operations/puppet
Gerrit-Branch: production
Gerrit-Owner: Ottomata <[email protected]>
_______________________________________________
MediaWiki-commits mailing list
[email protected]
https://lists.wikimedia.org/mailman/listinfo/mediawiki-commits
