file operations for reading and writing lists of nodes/sites/pcus, etc.

[monitor.git] / automate_pl03.sh

diff --git a/automate_pl03.sh b/automate_pl03.sh
index eccbc68..182aeec 100755 (executable)
--- a/automate_pl03.sh
+++ b/automate_pl03.sh
@@ -1,11 +1,13 @@
 #!/bin/bash
 
-source monitorconfig.py
+# NOTE: Must be an absolute path to guarantee it is read.
+source /home/soltesz/monitor/monitorconfig.py
 cd ${MONITOR_SCRIPT_ROOT}
 set -e
 DATE=`date +%Y-%m-%d-%T`
 MONITOR_PID="$HOME/monitor/SKIP"
 
+echo "Performing API test"
 API=$(./testapi.py)
 if [ "$API" != "ok" ] ; then 
        # NOTE: Do not try to run any commands if the API is obviously broken.
@@ -27,6 +29,7 @@ if [ -f $MONITOR_PID ] ; then
 fi
 echo $$ > $MONITOR_PID
 
+echo "Performing Findbad Nodes"
 #########################
 # 1. FINDBAD NODES 
 rm -f ${MONITOR_DATA_ROOT}/production.findbad2.pkl
@@ -34,6 +37,7 @@ ${MONITOR_SCRIPT_ROOT}/findbad.py --increment --cachenodes --debug=0 --dbname="f
 cp ${MONITOR_DATA_ROOT}/production.findbad2.pkl ${MONITOR_DATA_ROOT}/production.findbad.pkl
 ps ax | grep BatchMode | grep -v grep | awk '{print $1}' | xargs kill || :
 
+echo "Performing Findbad PCUs"
 #########################
 # 2. FINDBAD PCUS
 rm -f ${MONITOR_DATA_ROOT}/production.findbadpcus2.pkl
@@ -42,17 +46,20 @@ cp ${MONITOR_DATA_ROOT}/production.findbadpcus2.pkl ${MONITOR_DATA_ROOT}/product
 # clean up stray 'locfg' processes that hang around inappropriately...
 ps ax | grep locfg | grep -v grep | awk '{print $1}' | xargs kill || :
 
+echo "Generating web data"
 # badcsv.txt
 ${MONITOR_SCRIPT_ROOT}/printbadcsv.py  | grep -v loading | tr -d ' ' > badcsv.txt
 cp badcsv.txt /plc/data/var/www/html/monitor/
 ${MONITOR_SCRIPT_ROOT}/showlatlon.py | head -9 | awk 'BEGIN {print "<table>"} { print "<tr><td>", $0, "</td></tr>"} END{print "</table>"}'  | sed -e 's\|\</td><td>\g' > /plc/data/var/www/html/monitor/regions.html
 
+echo "Performing uptime changes for sites, nodes, and pcus"
 ########################
 # 3. record last-changed for sites, nodes and pcus.
 ${MONITOR_SCRIPT_ROOT}/sitebad.py --increment || :
 ${MONITOR_SCRIPT_ROOT}/nodebad.py --increment || :
 ${MONITOR_SCRIPT_ROOT}/pcubad.py --increment || :
 
+echo "Converting pkl files to phpserial"
 #########################
 # 4. convert pkl to php serialize format.
 ${MONITOR_SCRIPT_ROOT}/pkl2php.py -i findbadpcus2 -o findbadpcus
@@ -62,20 +69,23 @@ ${MONITOR_SCRIPT_ROOT}/pkl2php.py -i findbad -o findbadnodes
 ${MONITOR_SCRIPT_ROOT}/pkl2php.py -i ad_dbTickets -o ad_dbTickets
 ${MONITOR_SCRIPT_ROOT}/pkl2php.py -i idTickets -o idTickets
 
+echo "Archiving pkl files"
 #########################
 # Archive pkl files.
 for f in findbad act_all findbadpcus l_plcnodes site_persistflags node_persistflags pcu_persistflags ; do 
        cp ${MONITOR_DATA_ROOT}/production.$f.pkl ${MONITOR_ARCHIVE_ROOT}/`date +%F-%H:%M`.production.$f.pkl
 done
 
+echo "Running grouprins on all dbg nodes"
 ############################
 # 5. Check if there are any nodes in dbg state.  Clean up afterward.
 ${MONITOR_SCRIPT_ROOT}/grouprins.py --mail=1 \
-       --nodeselect 'state=DEBUG&&boot_state=dbg||state=DEBUG&&boot_state=boot' \
+       --nodeselect 'state=DEBUG&&boot_state=(rins|dbg|boot)' \
        --stopselect 'state=BOOT&&kernel=2.6.22.19-vs2.3.0.34.9.planetlab' \
        --reboot || :
 ${MONITOR_SCRIPT_ROOT}/findbad.py --increment --cachenodes --debug=0 --dbname="findbad" --nodeselect 'state=DEBUG&&boot_state=dbg||state=DEBUG&&boot_state=boot' || :
 
+echo "Collecting RT database dump"
 ##########################
 # 6. cache the RT db locally.
 python ${MONITOR_SCRIPT_ROOT}/rt.py