X-Git-Url: http://git.onelab.eu/?a=blobdiff_plain;f=automate-default.sh;h=046c1acb85d1047435d8bcf7a05e8ed31e34aab4;hb=38d434f7f399272701bf0d85fcb0f82ae3b44733;hp=8e7be9c5402921c91377e599ca332a1c6c807159;hpb=6f2351e4b44590221425fa9b4bfa77c92db49b6a;p=monitor.git diff --git a/automate-default.sh b/automate-default.sh index 8e7be9c..046c1ac 100755 --- a/automate-default.sh +++ b/automate-default.sh @@ -1,15 +1,16 @@ #!/bin/bash # NOTE: Must be an absolute path to guarantee it is read. -INSTALLPATH=/usr/share/monitor-server/ +INSTALLPATH=/usr/share/monitor/ # Generate an 'sh' style file full of variables in monitor.conf $INSTALLPATH/shconfig.py > $INSTALLPATH/monitorconfig.sh source $INSTALLPATH/monitorconfig.sh cd ${MONITOR_SCRIPT_ROOT} set -e DATE=`date +%Y-%m-%d-%T` -MONITOR_PID="$HOME/monitor/SKIP" +MONITOR_PID="${MONITOR_SCRIPT_ROOT}/SKIP" +echo "#######################################"; echo "Running Monitor at $DATE"; echo "######################################" echo "Performing API test" API=$(./testapi.py) if [ "$API" != "ok" ] ; then @@ -23,7 +24,12 @@ if [ -f $MONITOR_PID ] ; then echo "KILLING Monitor" PID=`cat $MONITOR_PID` rm -f $MONITOR_PID - ${MONITOR_SCRIPT_ROOT}/kill.cmd.sh $PID + if [ -z $PID ] ; then + ${MONITOR_SCRIPT_ROOT}/kill.cmd.sh $PID + echo "done." + else + echo "No PID to be killed." + fi else # skipping monitor echo "SKIPPING Monitor" @@ -35,13 +41,14 @@ echo $$ > $MONITOR_PID # SETUP act_all database if it's not there. if [ ! -f ${MONITOR_SCRIPT_ROOT}/actallsetup.flag ]; then if ! python -c 'import database; database.dbLoad("act_all")' 2>/dev/null ; then - python -c 'import database; database.dbDump("act_all", {})' 2>/dev/null ; then touch ${MONITOR_SCRIPT_ROOT}/actallsetup.flag fi fi +set +e AGENT=`ps ax | grep ssh-agent | grep -v grep` +set -e if [ -z "$AGENT" ] ; then echo "starting ssh agent" # if no agent is running, set it up. @@ -57,62 +64,39 @@ source ${MONITOR_SCRIPT_ROOT}/agent.sh echo "Performing Findbad Nodes" ######################### # 1. FINDBAD NODES -rm -f ${MONITOR_DATA_ROOT}/production.findbad2.pkl -${MONITOR_SCRIPT_ROOT}/findbad.py --increment --cachenodes --debug=0 --dbname="findbad2" $DATE || : -cp ${MONITOR_DATA_ROOT}/production.findbad2.pkl ${MONITOR_DATA_ROOT}/production.findbad.pkl -ps ax | grep BatchMode | grep -v grep | awk '{print $1}' | xargs kill || : +${MONITOR_SCRIPT_ROOT}/findbad.py --increment $DATE || : +ps ax | grep BatchMode | grep -v grep | awk '{print $1}' | xargs -r kill || : echo "Performing Findbad PCUs" ######################### # 2. FINDBAD PCUS -rm -f ${MONITOR_DATA_ROOT}/production.findbadpcus2.pkl -${MONITOR_SCRIPT_ROOT}/findbadpcu.py --increment --refresh --debug=0 --dbname=findbadpcus2 $DATE || : -cp ${MONITOR_DATA_ROOT}/production.findbadpcus2.pkl ${MONITOR_DATA_ROOT}/production.findbadpcus.pkl +${MONITOR_SCRIPT_ROOT}/findbadpcu.py --increment $DATE || : # 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 ""} { print ""} END{print "
", $0, "
"}' | sed -e 's\|\\g' > /plc/data/var/www/html/monitor/regions.html +ps ax | grep locfg | grep -v grep | awk '{print $1}' | xargs -r kill || : echo "Performing uptime changes for sites, nodes, and pcus" ######################## # 3. record last-changed for sites, nodes and pcus. -${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. -${MONITOR_SCRIPT_ROOT}/pkl2php.py -i findbadpcus2 -o findbadpcus -${MONITOR_SCRIPT_ROOT}/pkl2php.py -i act_all -o act_all -${MONITOR_SCRIPT_ROOT}/pkl2php.py -i plcdb_hn2lb -o plcdb_hn2lb -${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 +${MONITOR_SCRIPT_ROOT}/sitebad.py || : +${MONITOR_SCRIPT_ROOT}/nodebad.py || : +${MONITOR_SCRIPT_ROOT}/pcubad.py || : 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 +for f in findbad act_all findbadpcus l_plcnodes site_persistflags node_persistflags pcu_persistflags ; do + if [ -f ${MONITOR_DATA_ROOT}/production.$f.pkl ] ; then + cp ${MONITOR_DATA_ROOT}/production.$f.pkl ${MONITOR_ARCHIVE_ROOT}/`date +%F-%H:%M`.production.$f.pkl + else + echo "Warning: It failed to archive ${MONITOR_DATA_ROOT}/production.$f.pkl" + fi done -echo "Running grouprins on all dbg nodes" +#echo "Running grouprins on all dbg nodes" ############################ # 5. Check if there are any nodes in dbg state. Clean up afterward. -${MONITOR_SCRIPT_ROOT}/grouprins.py --mail=1 \ - --nodeselect 'state=DEBUG&&boot_state=(rins|dbg|boot)' \ - --stopselect 'state=BOOT&&kernel=2.6.22.19-vs2.3.0.34.9.planetlab' \ - --reboot || : -${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 ${MONITOR_SCRIPT_ROOT}/rt.py +#${MONITOR_SCRIPT_ROOT}/grouprins.py --mail=1 --reboot --nodeselect 'state=DOWN&&boot_state=(boot|rins|dbg|diag)' --stopselect "state=BOOT" || : +#${MONITOR_SCRIPT_ROOT}/grouprins.py --mail=1 --reboot --nodeselect 'state=DEBUG&&boot_state=(rins|dbg|boot)' --stopselect 'state=BOOT' || : +cp ${MONITOR_SCRIPT_ROOT}/monitor.log ${MONITOR_ARCHIVE_ROOT}/`date +%F-%H:%M`.monitor.log rm -f $MONITOR_PID