From: James Nunez Date: Thu, 6 Sep 2018 20:34:36 +0000 (-0600) Subject: LU-11349 tests: remove insanity test 15 X-Git-Tag: 2.11.56~77 X-Git-Url: https://git.whamcloud.com/?p=fs%2Flustre-release.git;a=commitdiff_plain;h=7b2dcd2c348d1b3da59095260edbd31383baea30;hp=48e409e65edda13dd647cf37458a7a8ae8218a23 LU-11349 tests: remove insanity test 15 Test 15 of the insanity test suite has not been used in testing since 2003 and tries to run a script that no longer exists. Thus, this test needs to be removed. At the same time, clean up the script in converting spaces to tabs. Test-Parameters: trivial testlist=insanity Signed-off-by: James Nunez Change-Id: Ie3b2f88dde925d7a1c53f288215b3643d20fd5ff Reviewed-on: https://review.whamcloud.com/33121 Tested-by: Jenkins Tested-by: Maloo Reviewed-by: Emoly Liu Reviewed-by: Wei Liu Reviewed-by: Oleg Drokin --- diff --git a/lustre/tests/insanity.sh b/lustre/tests/insanity.sh index 37f2f66..1d91a59 100755 --- a/lustre/tests/insanity.sh +++ b/lustre/tests/insanity.sh @@ -1,8 +1,6 @@ #!/bin/bash -# -*- mode: Bash; tab-width: 4; indent-tabs-mode: t; -*- -# vim:shiftwidth=4:softtabstop=4:tabstop=4: # -# Test multiple failures, AKA Test 17 +# Test different failure modes combinations set -e @@ -13,8 +11,10 @@ init_test_env $@ . ${CONFIG:=$LUSTRE/tests/cfg/$NAME.sh} init_logging -# -ALWAYS_EXCEPT="15 $INSANITY_EXCEPT" + +# bug number for skipped test: +ALWAYS_EXCEPT="$INSANITY_EXCEPT" +# UPDATE THE COMMENT ABOVE WITH BUG NUMBERS WHEN CHANGING ALWAYS_EXCEPT! if [ "$FAILURE_MODE" = "HARD" ]; then skip_env "$TESTSUITE: is not functional with FAILURE_MODE = HARD, " \ @@ -57,9 +57,9 @@ DOWN_NUM=0 # number of nodes currently down # set next client to fail set_fail_client() { - FAIL_CLIENT=${FAIL_LIST[$FAIL_NEXT]} - FAIL_NEXT=$(( (FAIL_NEXT+1) % FAIL_NUM )) - echo "fail $FAIL_CLIENT, next is $FAIL_NEXT" + FAIL_CLIENT=${FAIL_LIST[$FAIL_NEXT]} + FAIL_NEXT=$(( (FAIL_NEXT+1) % FAIL_NUM )) + echo "fail $FAIL_CLIENT, next is $FAIL_NEXT" } fail_clients() { @@ -69,22 +69,22 @@ fail_clients() { if [ -z "$num" ] || [ "$num" -gt $((FAIL_NUM - DOWN_NUM)) ]; then num=$((FAIL_NUM - DOWN_NUM)) fi - - if [ -z "$num" ] || [ "$num" -le 0 ]; then - log "No clients failed!" - return - fi - client_mkdirs + if [ -z "$num" ] || [ "$num" -le 0 ]; then + log "No clients failed!" + return + fi + + client_mkdirs - for i in `seq $num`; do - set_fail_client - client=$FAIL_CLIENT - DOWN_CLIENTS="$DOWN_CLIENTS $client" - shutdown_client $client - done + for i in `seq $num`; do + set_fail_client + client=$FAIL_CLIENT + DOWN_CLIENTS="$DOWN_CLIENTS $client" + shutdown_client $client + done - echo "down clients: $DOWN_CLIENTS" + echo "down clients: $DOWN_CLIENTS" for client in $DOWN_CLIENTS; do boot_node $client @@ -153,18 +153,15 @@ check_and_setup_lustre rm -rf $TESTDIR mkdir -p $TESTDIR -# 9 Different Failure Modes Combinations -echo "Starting Test 17 at `date`" - test_0() { for i in $(seq $MDSCOUNT) ; do fail mds$i done - for i in $(seq $OSTCOUNT) ; do - fail ost$i - done - return 0 + for i in $(seq $OSTCOUNT) ; do + fail ost$i + done + return 0 } run_test 0 "Fail all nodes, independently" @@ -219,27 +216,27 @@ test_2() { reboot_facet mds$i done - clients_up & - DFPID=$! - sleep 5 + clients_up & + DFPID=$! + sleep 5 - shutdown_facet ost1 + shutdown_facet ost1 - echo "Reintegrating OST" - reboot_facet ost1 - wait_for_facet ost1 - start_ost 1 || return 2 + echo "Reintegrating OST" + reboot_facet ost1 + wait_for_facet ost1 + start_ost 1 || return 2 for i in $(seq $MDSCOUNT) ; do wait_for_facet mds$i start_mdt $i || return $? done - #Check FS - wait $DFPID - clients_recover_osts ost1 - echo "Verify reintegration" - clients_up || return 1 + #Check FS + wait $DFPID + clients_recover_osts ost1 + echo "Verify reintegration" + clients_up || return 1 } run_test 2 "Second Failure Mode: MDS/OST `date`" @@ -247,52 +244,52 @@ run_test 2 "Second Failure Mode: MDS/OST `date`" ############### Third Failure Mode ############### test_3() { - #Create files - echo "Verify Lustre filesystem is up and running" - [ -z "$(mounted_lustre_filesystems)" ] && error "Lustre is not running" + #Create files + echo "Verify Lustre filesystem is up and running" + [ -z "$(mounted_lustre_filesystems)" ] && error "Lustre is not running" - #MDS Portion + #MDS Portion for i in $(seq $MDSCOUNT) ; do fail mds$i done - #Check FS - - echo "Test Lustre stability after MDS failover" - clients_up - - #CLIENT Portion - echo "Failing 2 CLIENTS" - fail_clients 2 - - #Check FS - echo "Test Lustre stability after CLIENT failure" - clients_up - - #Reintegration - echo "Reintegrating CLIENTS" - reintegrate_clients || return 1 - - clients_up || return 3 - sleep 2 # give it a little time for fully recovered before next test + + #Check FS + echo "Test Lustre stability after MDS failover" + clients_up + + #CLIENT Portion + echo "Failing 2 CLIENTS" + fail_clients 2 + + #Check FS + echo "Test Lustre stability after CLIENT failure" + clients_up + + #Reintegration + echo "Reintegrating CLIENTS" + reintegrate_clients || return 1 + + clients_up || return 3 + sleep 2 # give it a little time for fully recovered before next test } -run_test 3 "Thirdb Failure Mode: MDS/CLIENT `date`" +run_test 3 "Third Failure Mode: MDS/CLIENT `date`" ################################################### ############### Fourth Failure Mode ############### test_4() { echo "Fourth Failure Mode: OST/MDS `date`" - #OST Portion - shutdown_facet ost1 + #OST Portion + shutdown_facet ost1 - #Check FS - echo "Test Lustre stability after OST failure" - clients_up & - DFPIDA=$! - sleep 5 + #Check FS + echo "Test Lustre stability after OST failure" + clients_up & + DFPIDA=$! + sleep 5 for i in $(seq $MDSCOUNT) ; do - shutdown_facet mds$i + shutdown_facet mds$i reboot_facet mds$i # prepare for MDS failover @@ -300,27 +297,27 @@ test_4() { reboot_facet mds$i done - clients_up & - DFPIDB=$! - sleep 5 + clients_up & + DFPIDB=$! + sleep 5 - #Reintegration - echo "Reintegrating OST" - reboot_facet ost1 - wait_for_facet ost1 - start_ost 1 + #Reintegration + echo "Reintegrating OST" + reboot_facet ost1 + wait_for_facet ost1 + start_ost 1 for i in $(seq $MDSCOUNT) ; do wait_for_facet mds$i start_mdt $i || return $? done - #Check FS - wait $DFPIDA - wait $DFPIDB - clients_recover_osts ost1 - echo "Test Lustre stability after MDS failover" - clients_up || return 1 + #Check FS + wait $DFPIDA + wait $DFPIDB + clients_recover_osts ost1 + echo "Test Lustre stability after MDS failover" + clients_up || return 1 } run_test 4 "Fourth Failure Mode: OST/MDS `date`" ################################################### @@ -340,256 +337,250 @@ test_5() { shutdown_facet ost1 reboot_facet ost1 - #Check FS - echo "Test Lustre stability after OST failure" - clients_up & - DFPIDA=$! - sleep 5 - - #OST Portion - shutdown_facet ost2 - reboot_facet ost2 - - #Check FS - echo "Test Lustre stability after OST failure" - clients_up & - DFPIDB=$! - sleep 5 - - #Reintegration - echo "Reintegrating OSTs" - wait_for_facet ost1 - start_ost 1 - wait_for_facet ost2 - start_ost 2 - - clients_recover_osts ost1 - clients_recover_osts ost2 - sleep $TIMEOUT - - wait $DFPIDA - wait $DFPIDB - clients_up || return 2 + #Check FS + echo "Test Lustre stability after OST failure" + clients_up & + DFPIDA=$! + sleep 5 + + #OST Portion + shutdown_facet ost2 + reboot_facet ost2 + + #Check FS + echo "Test Lustre stability after OST failure" + clients_up & + DFPIDB=$! + sleep 5 + + #Reintegration + echo "Reintegrating OSTs" + wait_for_facet ost1 + start_ost 1 + wait_for_facet ost2 + start_ost 2 + + clients_recover_osts ost1 + clients_recover_osts ost2 + sleep $TIMEOUT + + wait $DFPIDA + wait $DFPIDB + clients_up || return 2 } run_test 5 "Fifth Failure Mode: OST/OST `date`" ################################################### ############### Sixth Failure Mode ############### test_6() { - echo "Sixth Failure Mode: OST/CLIENT `date`" - - #Create files - echo "Verify Lustre filesystem is up and running" - [ -z "$(mounted_lustre_filesystems)" ] && error "Lustre is not running" - - clients_up - client_touch testfile || return 2 - - #OST Portion - shutdown_facet ost1 - reboot_facet ost1 - - #Check FS - echo "Test Lustre stability after OST failure" - clients_up & - DFPIDA=$! - echo DFPIDA=$DFPIDA - sleep 5 - - #CLIENT Portion - echo "Failing CLIENTs" - fail_clients - - #Check FS - echo "Test Lustre stability after CLIENTs failure" - clients_up & - DFPIDB=$! - echo DFPIDB=$DFPIDB - sleep 5 - - #Reintegration - echo "Reintegrating OST/CLIENTs" - wait_for_facet ost1 - start_ost 1 - reintegrate_clients || return 1 - sleep 5 - - wait_remote_prog "stat -f" $((TIMEOUT * 3 + 20)) - wait $DFPIDA - wait $DFPIDB - - echo "Verifying mount" - [ -z "$(mounted_lustre_filesystems)" ] && return 3 - clients_up + echo "Sixth Failure Mode: OST/CLIENT `date`" + + #Create files + echo "Verify Lustre filesystem is up and running" + [ -z "$(mounted_lustre_filesystems)" ] && error "Lustre is not running" + + clients_up + client_touch testfile || return 2 + + #OST Portion + shutdown_facet ost1 + reboot_facet ost1 + + #Check FS + echo "Test Lustre stability after OST failure" + clients_up & + DFPIDA=$! + echo DFPIDA=$DFPIDA + sleep 5 + + #CLIENT Portion + echo "Failing CLIENTs" + fail_clients + + #Check FS + echo "Test Lustre stability after CLIENTs failure" + clients_up & + DFPIDB=$! + echo DFPIDB=$DFPIDB + sleep 5 + + #Reintegration + echo "Reintegrating OST/CLIENTs" + wait_for_facet ost1 + start_ost 1 + reintegrate_clients || return 1 + sleep 5 + + wait_remote_prog "stat -f" $((TIMEOUT * 3 + 20)) + wait $DFPIDA + wait $DFPIDB + + echo "Verifying mount" + [ -z "$(mounted_lustre_filesystems)" ] && return 3 + clients_up } run_test 6 "Sixth Failure Mode: OST/CLIENT `date`" ################################################### - ############### Seventh Failure Mode ############### test_7() { - echo "Seventh Failure Mode: CLIENT/MDS `date`" - - #Create files - echo "Verify Lustre filesystem is up and running" - [ -z "$(mounted_lustre_filesystems)" ] && error "Lustre is not running" - - clients_up - client_touch testfile || return 1 - - #CLIENT Portion - echo "Part 1: Failing CLIENT" - fail_clients 2 - - #Check FS - echo "Test Lustre stability after CLIENTs failure" - clients_up - $PDSH $LIVE_CLIENT "ls -l $TESTDIR" - $PDSH $LIVE_CLIENT "rm -f $TESTDIR/*_testfile" - - #Sleep - echo "Wait 1 minutes" - sleep 60 - - #Create files - echo "Verify Lustre filesystem is up and running" - [ -z "$(mounted_lustre_filesystems)" ] && return 2 - - clients_up - client_rm testfile - - #MDS Portion + echo "Seventh Failure Mode: CLIENT/MDS `date`" + + #Create files + echo "Verify Lustre filesystem is up and running" + [ -z "$(mounted_lustre_filesystems)" ] && error "Lustre is not running" + + clients_up + client_touch testfile || return 1 + + #CLIENT Portion + echo "Part 1: Failing CLIENT" + fail_clients 2 + + #Check FS + echo "Test Lustre stability after CLIENTs failure" + clients_up + $PDSH $LIVE_CLIENT "ls -l $TESTDIR" + $PDSH $LIVE_CLIENT "rm -f $TESTDIR/*_testfile" + + #Sleep + echo "Wait 1 minutes" + sleep 60 + + #Create files + echo "Verify Lustre filesystem is up and running" + [ -z "$(mounted_lustre_filesystems)" ] && return 2 + + clients_up + client_rm testfile + + #MDS Portion for i in $(seq $MDSCOUNT) ; do fail mds$i done - $PDSH $LIVE_CLIENT "ls -l $TESTDIR" - $PDSH $LIVE_CLIENT "rm -f $TESTDIR/*_testfile" - - #Reintegration - echo "Reintegrating CLIENTs" - reintegrate_clients || return 2 - clients_up - - #Sleep - echo "wait 1 minutes" - sleep 60 + $PDSH $LIVE_CLIENT "ls -l $TESTDIR" + $PDSH $LIVE_CLIENT "rm -f $TESTDIR/*_testfile" + + #Reintegration + echo "Reintegrating CLIENTs" + reintegrate_clients || return 2 + clients_up + + #Sleep + echo "wait 1 minutes" + sleep 60 } run_test 7 "Seventh Failure Mode: CLIENT/MDS `date`" ################################################### - ############### Eighth Failure Mode ############### test_8() { - echo "Eighth Failure Mode: CLIENT/OST `date`" - - #Create files - echo "Verify Lustre filesystem is up and running" - [ -z "$(mounted_lustre_filesystems)" ] && error "Lustre is not running" - - clients_up - client_touch testfile - - #CLIENT Portion - echo "Failing CLIENTs" - fail_clients 2 - - #Check FS - echo "Test Lustre stability after CLIENTs failure" - clients_up - $PDSH $LIVE_CLIENT "ls -l $TESTDIR" - $PDSH $LIVE_CLIENT "rm -f $TESTDIR/*_testfile" - - #Sleep - echo "Wait 1 minutes" - sleep 60 - - #Create files - echo "Verify Lustre filesystem is up and running" - [ -z "$(mounted_lustre_filesystems)" ] && error "Lustre is not running" - - clients_up - client_touch testfile - - - #OST Portion - shutdown_facet ost1 - reboot_facet ost1 - - #Check FS - echo "Test Lustre stability after OST failure" - clients_up & - DFPID=$! - sleep 5 - #non-failout hangs forever here - #$PDSH $LIVE_CLIENT "ls -l $TESTDIR" - #$PDSH $LIVE_CLIENT "rm -f $TESTDIR/*_testfile" - - #Reintegration - echo "Reintegrating CLIENTs/OST" - reintegrate_clients || return 3 - wait_for_facet ost1 - start_ost 1 - wait $DFPID - clients_up || return 1 - client_touch testfile2 || return 2 - - #Sleep - echo "Wait 1 minutes" - sleep 60 + echo "Eighth Failure Mode: CLIENT/OST `date`" + + #Create files + echo "Verify Lustre filesystem is up and running" + [ -z "$(mounted_lustre_filesystems)" ] && error "Lustre is not running" + + clients_up + client_touch testfile + + #CLIENT Portion + echo "Failing CLIENTs" + fail_clients 2 + + #Check FS + echo "Test Lustre stability after CLIENTs failure" + clients_up + $PDSH $LIVE_CLIENT "ls -l $TESTDIR" + $PDSH $LIVE_CLIENT "rm -f $TESTDIR/*_testfile" + + #Sleep + echo "Wait 1 minutes" + sleep 60 + + #Create files + echo "Verify Lustre filesystem is up and running" + [ -z "$(mounted_lustre_filesystems)" ] && error "Lustre is not running" + + clients_up + client_touch testfile + + #OST Portion + shutdown_facet ost1 + reboot_facet ost1 + + #Check FS + echo "Test Lustre stability after OST failure" + clients_up & + DFPID=$! + sleep 5 + #non-failout hangs forever here + #$PDSH $LIVE_CLIENT "ls -l $TESTDIR" + #$PDSH $LIVE_CLIENT "rm -f $TESTDIR/*_testfile" + + #Reintegration + echo "Reintegrating CLIENTs/OST" + reintegrate_clients || return 3 + wait_for_facet ost1 + start_ost 1 + wait $DFPID + clients_up || return 1 + client_touch testfile2 || return 2 + + #Sleep + echo "Wait 1 minutes" + sleep 60 } run_test 8 "Eighth Failure Mode: CLIENT/OST `date`" ################################################### - ############### Ninth Failure Mode ############### test_9() { - echo - - #Create files - echo "Verify Lustre filesystem is up and running" - [ -z "$(mounted_lustre_filesystems)" ] && error "Lustre is not running" - - clients_up - client_touch testfile || return 1 - - #CLIENT Portion - echo "Failing CLIENTs" - fail_clients 2 - - #Check FS - echo "Test Lustre stability after CLIENTs failure" - clients_up - $PDSH $LIVE_CLIENT "ls -l $TESTDIR" || return 1 - $PDSH $LIVE_CLIENT "rm -f $TESTDIR/*_testfile" || return 2 - - #Sleep - echo "Wait 1 minutes" - sleep 60 - - #Create files - echo "Verify Lustre filesystem is up and running" - client_up $LIVE_CLIENT || return 3 - client_touch testfile || return 4 - - #CLIENT Portion - echo "Failing CLIENTs" - fail_clients 2 - - #Check FS - echo "Test Lustre stability after CLIENTs failure" - clients_up - $PDSH $LIVE_CLIENT "ls -l $TESTDIR" || return 5 - $PDSH $LIVE_CLIENT "rm -f $TESTDIR/*_testfile" || return 6 - - #Reintegration - echo "Reintegrating CLIENTs/CLIENTs" - reintegrate_clients || return 7 - clients_up - - #Sleep - echo "Wait 1 minutes" - sleep 60 + #Create files + echo "Verify Lustre filesystem is up and running" + [ -z "$(mounted_lustre_filesystems)" ] && error "Lustre is not running" + + clients_up + client_touch testfile || return 1 + + #CLIENT Portion + echo "Failing CLIENTs" + fail_clients 2 + + #Check FS + echo "Test Lustre stability after CLIENTs failure" + clients_up + $PDSH $LIVE_CLIENT "ls -l $TESTDIR" || return 1 + $PDSH $LIVE_CLIENT "rm -f $TESTDIR/*_testfile" || return 2 + + #Sleep + echo "Wait 1 minutes" + sleep 60 + + #Create files + echo "Verify Lustre filesystem is up and running" + client_up $LIVE_CLIENT || return 3 + client_touch testfile || return 4 + + #CLIENT Portion + echo "Failing CLIENTs" + fail_clients 2 + + #Check FS + echo "Test Lustre stability after CLIENTs failure" + clients_up + $PDSH $LIVE_CLIENT "ls -l $TESTDIR" || return 5 + $PDSH $LIVE_CLIENT "rm -f $TESTDIR/*_testfile" || return 6 + + #Reintegration + echo "Reintegrating CLIENTs/CLIENTs" + reintegrate_clients || return 7 + clients_up + + #Sleep + echo "Wait 1 minutes" + sleep 60 } run_test 9 "Ninth Failure Mode: CLIENT/CLIENT `date`" ################################################### @@ -762,15 +753,6 @@ test_14() { clients_up || return 4 } run_test 14 "Fourteen Failure Mode: OST0,OST1/CLIENTS/MDS0,MDS1 `date`" -################################################### - -test_15() { - #Run availability after all failures - DURATION=${DURATION:-$((2 * 60 * 60))} # 6 hours default - LOADTEST=${LOADTEST:-metadata-load.py} - $PWD/availability.sh $CONFIG $DURATION $CLIENTS || return 1 -} -run_test 15 "Running Availability for 6 hours..." complete $SECONDS check_and_cleanup_lustre