Document how to fix BASE-IS-OLD

[mirror/dsa-wiki.git] / input / howto / postgres-backup.creole

diff --git a/input/howto/postgres-backup.creole b/input/howto/postgres-backup.creole
index d6957aa..fee97c6 100644 (file)
--- a/input/howto/postgres-backup.creole
+++ b/input/howto/postgres-backup.creole
@@ -2,7 +2,7 @@
 
 Backing up postgres consists of two main pieces: backing up the WALs (write
 ahead logs) and regular backups of the base.  See
-http://www.postgresql.org/docs/9.4/static/continuous-archiving.html
+https://www.postgresql.org/docs/9.4/static/continuous-archiving.html
 
 == WAL Shipping ==
 
@@ -12,7 +12,7 @@ http://www.postgresql.org/docs/9.4/static/continuous-archiving.html
 }}}
 * clone dsa-misc:
 {{{
-  cd /usr/local && git clone https://db.debian.org/git/dsa-misc.git &&
+  cd /usr/local && git -c http.sslCAInfo=/etc/ssl/ca-debian/ca-certificates.crt clone https://db.debian.org/git/dsa-misc.git &&
   cd bin && ln -s ../dsa-misc/scripts/pg-backup/pg* .
 }}}
 * Add these to /etc/postgresql/9.1/main/postgresql.conf
@@ -50,3 +50,19 @@ http://www.postgresql.org/docs/9.4/static/continuous-archiving.html
 
 * Test running "postgres-make-base-backups host:port".
 * You should see a tarball and WALs
+
+= Nagios warnings =
+
+== BASE-IS-OLD ==
+
+(2018-02) Our nagios check warns us when a backup server has not successfully fetched
+a base backup recently.  The causes often are that either the postgres server or the
+backup host went down or was down during the time of the weekly cronjob.
+
+To re-run a base backup for a specific cluster, log into the backup server
+(either storace or backuphost), cat /usr/local/bin/postgres-make-base-backups
+to see the port for the cluster, and run
+{{{
+  sudo -u debbackup /usr/local/bin/postgres-make-base-backups <host>:<port>
+}}}
+probably best to do that in a screen as it might take a while.