github · rubiojr · Jun 21, 2016 · Jun 20, 2016
diff --git a/bin/ghe-restore b/bin/ghe-restore
@@ -217,7 +217,12 @@ bm_end "ghe-import-redis"
 
 if $cluster; then
   echo "Restoring Git repositories into cluster ..."
-  ghe-restore-repositories-dgit "$GHE_HOSTNAME" 1>&3
+  if ghe-ssh "$GHE_HOSTNAME" test -f /data/github/current/script/dgit-cluster-restore-routes; then
+    ghe_verbose "* Using ghe-restore-repositories-dgit-ng to restore"
+    ghe-restore-repositories-dgit-ng "$GHE_HOSTNAME" 1>&3
+  else
+    ghe-restore-repositories-dgit "$GHE_HOSTNAME" 1>&3
+  fi
 
   echo "Restoring Gists into cluster ..."
   if ghe-ssh "$GHE_HOSTNAME" test -f /data/github/current/script/gist-cluster-restore-routes; then

diff --git a/share/github-backup-utils/ghe-restore-repositories-dgit-ng b/share/github-backup-utils/ghe-restore-repositories-dgit-ng
@@ -0,0 +1,146 @@
+#!/usr/bin/env bash
+#/ Usage: ghe-restore-repositories-dgit-ng <host>
+#/ Restore repositories fron an rsync snapshot of all Git repository data to a GitHub cluster.
+#/
+#/ Note: This script typically isn't called directly. It's invoked by the
+#/ ghe-restore command when restoring into a cluster.
+set -e
+
+# Bring in the backup configuration
+. $( dirname "${BASH_SOURCE[0]}" )/ghe-backup-config
+
+# Show usage and bail with no arguments
+[ -z "$*" ] && print_usage
+
+bm_start "$(basename $0)"
+
+# Grab host arg
+GHE_HOSTNAME="$1"
+
+# The snapshot to restore should be set by the ghe-restore command but this lets
+# us run this script directly.
+: ${GHE_RESTORE_SNAPSHOT:=current}
+
+network_paths=$(cd $GHE_DATA_DIR/$GHE_RESTORE_SNAPSHOT/ && find repositories -mindepth 6 -maxdepth 7 -name \*.git -exec dirname {} \; | uniq | grep nw | cut -d / -f2-)
+
+if [ -z "$network_paths" ]; then
+  echo "Warning: Repositories backup missing. Skipping ..."
+  exit 0
+fi
+
+# Perform a host-check and establish GHE_REMOTE_XXX variables.
+ghe_remote_version_required "$GHE_HOSTNAME"
+
+# Generate SSH config for forwarding
+# Split host:port into parts
+port=$(ssh_port_part "$GHE_HOSTNAME")
+host=$(ssh_host_part "$GHE_HOSTNAME")
+
+# Add user / -l option
+user="${host%@*}"
+[ "$user" = "$host" ] && user="admin"
+
+tempdir=$(mktemp -d)
+ssh_config_file=$tempdir/ssh_config
+opts="$GHE_EXTRA_SSH_OPTS -o UserKnownHostsFile=/dev/null -o StrictHostKeyChecking=no -o PasswordAuthentication=no"
+tmp_list=$tempdir/tmp_list
+to_restore=$tempdir/to_restore
+
+hostnames=$(ghe-ssh "$GHE_HOSTNAME" ghe-config --get-regexp cluster.*.hostname | cut -d ' ' -f 2)
+for hostname in $hostnames; do
+  echo "
+Host $hostname
+  ServerAliveInterval 60
+  ProxyCommand ssh -q $GHE_EXTRA_SSH_OPTS -p $port $user@$host nc.openbsd %h %p
+  StrictHostKeyChecking=no" >> $ssh_config_file
+done
+
+cleanup() {
+  for hostname in $hostnames; do
+    ghe-gc-enable -F $ssh_config_file $hostname:$port
+  done
+  rm -rf $tempdir
+}
+trap cleanup EXIT
+
+# Disable remote GC operations
+for hostname in $hostnames; do
+  ghe-gc-disable -F $ssh_config_file $hostname:$port
+done
+
+# Build a list of network paths to send to the server to calculate
+# the restore routes, something like:
+#
+# a/nw/a8/3f/02/100000855
+# a/nw/a8/bc/8d/100000880
+# a/nw/a5/06/81/100000659
+# a/nw/a5/84/6f/100000708
+# a/nw/a5/e0/01/146
+# ...
+#
+# One network path per line.
+OLDIFS=$IFS; IFS=$'\n'
+for path in $network_paths; do
+   # Get the network ID
+   # The nework id from a repository is the last component of the path
+   # i.e. /data/repositories/a/nw/a5/bf/c9/37 network ID would be 37
+   ghe_verbose "Adding network_path $path to the list of networks to send"
+   echo $path
+done > $tmp_list
+IFS=$OLDIFS
+
+# The server returns a list of routes:
+#
+# a/nw/a8/3f/02/100000855 dgit-node1 dgit-node2 dgit-node3
+# a/nw/a8/bc/8d/100000880 dgit-node1 dgit-node2 dgit-node4
+# a/nw/a5/06/81/100000659 dgit-node3 dgit-node2 dgit-node4
+# ...
+#
+# One route per line.
+cat $tmp_list | ghe-ssh "$GHE_HOSTNAME" github-env ./bin/dgit-cluster-restore-routes \
+ | while read route; do
+  ghe_verbose "Received route $route"
+  servers=$(echo $route | cut -d ' ' -f2-)
+  for server in $servers; do
+    network_path=$(echo $route | cut -d ' ' -f1)
+    ghe_verbose "Adding $network_path to $tempdir/$server.rsync"
+    echo "$network_path" >> $tempdir/$server.rsync
+  done
+
+  network_id=$(echo $network_path | awk -F/ '{print $(NF)}')
+  ghe_verbose "Route: $network_id /data/repositories/$network_path $servers"
+  echo "$network_id /data/repositories/$network_path $servers" >> $to_restore
+done
+
+# rsync all the repositories
+for file_list in $tempdir/*.rsync; do
+  server=$(basename $file_list .rsync)
+  ghe_verbose "* Transferring repositories to $server"
+  ghe-rsync -avrHR --delete \
+    -e "ssh -q $opts -p $port -F $ssh_config_file -l $user" \
+    --rsync-path="sudo -u git rsync" \
+    --files-from=$file_list \
+    "$GHE_DATA_DIR/$GHE_RESTORE_SNAPSHOT/repositories/./" \
+    "$server:$GHE_REMOTE_DATA_USER_DIR/repositories/" 1>&3
+done
+
+# Tell dgit about the repositories restored
+cat $to_restore | ghe-ssh "$GHE_HOSTNAME" github-env ./bin/dgit-cluster-restore-finalize >&3
+
+if [ -d $GHE_DATA_DIR/$GHE_RESTORE_SNAPSHOT/repositories/info ]; then
+  ghe_verbose "* Transferring repository info data"
+  for route in `ghe-ssh "$GHE_HOSTNAME" ghe-cluster-each -r git -p`; do
+    if ! ghe-rsync -av --delete \
+        -e "ssh -q $opts -p $port -F $ssh_config_file -l $user" \
+        --rsync-path="sudo -u git rsync" \
+        "$GHE_DATA_DIR/$GHE_RESTORE_SNAPSHOT/repositories/info/" \
+        "$route:$GHE_REMOTE_DATA_USER_DIR/repositories/info" 1>&3; then
+      echo "Error restoring /data/repositories/info to $route"
+    fi
+  done
+else
+  ghe_verbose "* Removing repository info data"
+  ghe-ssh "$GHE_HOSTNAME" ghe-cluster-each -r git -- rm -f /data/repositories/info/*
+fi
+
+bm_end "$(basename $0)"