#!/bin/bash #============================================================================== # Copyright and license info is available in the LICENSE file included with # the Server Deployment Package (SDP), and also available online: # https://swarm.workshop.perforce.com/projects/perforce-software-sdp/view/main/LICENSE #------------------------------------------------------------------------------ # verify_sdp.sh # Verifies SDP structure and environment. #============================================================================== # Declarations and Environment export VS_SDP_P4CBIN="/p4/common/bin" export VS_SDP_ENV="$VS_SDP_P4CBIN/p4_vars" export SDP_INSTANCE="${SDP_INSTANCE:-UnsetSDPInstance}" export VS_SDP_OWNER declare Version=5.13.0 declare SDPVersionA= declare SDPVersionB= declare SDPVersionC= declare ThisScript="${0##*/}" declare CmdArgs="$*" declare CmdLine="$0 $CmdArgs" declare -i ServerOnline=0 declare -i ErrorCount=0 declare -i WarningCount=0 declare -i CheckCount=0 declare -i SilentMode=0 declare -i DoCrontabTest=1 declare -i DoLicenseTest=1 declare -i DoVersionTest=1 declare -i DoExcessBinaryTest=1 declare -i DoOfflineDBTest=1 declare -i DoMasterIDTest=1 declare -i ExcessServerBinariesFound=0 declare -i P4DServer=0 declare -i P4BrokerServer=0 declare -i P4ProxyServer=0 declare SkipTestList= declare BadLog= declare ThisUser= declare LicenseInfo= declare LicenseExpiration= declare CurrentTime= declare ExpirationTime= declare TimeDiff= declare DaysDiff= declare LicenseDaysExpireWarning=14 declare LinkP4ROOT= declare LinkOfflineDB= declare H1="==============================================================================" declare H2="------------------------------------------------------------------------------" declare Log="Unset" export P4TMP="Unset" declare TmpFile= #============================================================================== # Local Functions # Note: This script does not use SDP library files, as its purpose is to # verify the integrity of an SDP installation. Thus, it has its own # self-contained versions of some functions for which similar versions # would normally be sourced from files in /p4/common/lib, like libbcore.sh. # Micro-functions, small or one-liners used to avoid external dependencies, # which is critical for this particular verify_sdp.sh script. function msg () { echo -e "$*"; } function errmsg () { msg "\\nError: ${1:-Unknown Error}\\n"; ErrorCount+=1; } function warnmsg () { msg "\\nWarning: ${1:-Unknown Warning}\\n"; WarningCount+=1; } function bail () { errmsg "${1:-Unknown Error}"; exit "${2:-1}" } #------------------------------------------------------------------------------ # This function takes as input an SDP version string, and returns a version # id of the form YYYY.N.CL, where YYYY is the year, N is an incrementing # release ID with a given year, and CL is a changelist identifier. The # YYYY.N together comprise the major version, often shortened to YY.N, e.g. # r20.1 for the 2020.1 release. # # The full SDP Version string looks something like this: # Rev. SDP/MultiArch/2019.3/26494 (2020/04/23). # # This function parses that full string and returns a value like: 2019.3.26494 function get_sdp_version_from_string () { local versionString="${1:-}" local version= version="20${versionString##*/20}" version="${version%% *}" version="${version/\//.}" [[ "$version" == "20" || "$version" == "200" ]] && version="Unknown" echo "$version" } #------------------------------------------------------------------------------ # Function: run ($cmd, $desc, $showOutput) # # Runs a command, with optional description, showing command line to execute # and optionally also the output, and capturing and returning the exit code. # # Input: # $1 - Command and arguments to execute. Defaults to 'echo'. # $2 - Optional message to display describing what the command is doing. # $3 - Numeric flag to show output; '1' indicates to show output, 0 to # suppress it. #------------------------------------------------------------------------------ function run () { local cmd="${1:-echo}" local desc="${2:-}" local -i showOutput="${3:-1}" local -i exitCode= local log log="$(mktemp "$P4TMP/run.XXXXXXXXXXX")" [[ -n "$desc" ]] && msg "$desc" msg "Executing: $cmd" $cmd > "$log" 2>&1 exitCode=$? if [[ "$showOutput" -eq 1 ]]; then echo "EXIT_CODE: $exitCode" >> "$log" cat "$log" fi /bin/rm -f "$log" return "$exitCode" } #------------------------------------------------------------------------------ # Function: usage (required function) # # Input: # $1 - style, either -h (for short form) or -man (for man-page like format). # The default is -h. # # $2 - error message (optional). Specify this if usage() is called due to # user error, in which case the given message displayed first, followed by the # standard usage message (short or long depending on $1). If displaying an # errror, usually $1 should be -h so that the longer usage message doesn't # obsure the error message. # # Sample Usage: # usage # usage -man # usage -h "Incorrect command line usage." # # This last example generates a usage error message followed by the short # '-h' usage summary. #------------------------------------------------------------------------------ function usage { declare style=${1:--h} declare errorMessage=${2:-Unset} if [[ $errorMessage != Unset ]]; then echo -e "\\n\\nUsage Error:\\n\\n$errorMessage\\n\\n" >&2 fi # tag::includeManual[] echo "USAGE for verify_sdp.sh v$Version: verify_sdp.sh [<instance>] [-online] [-skip <test>[,<test2>,...]] [-si] [-L <log>|off ] [-D] or verify_sdp.sh -h|-man " # end::includeManual[] if [[ $style == -man ]]; then # tag::includeManual[] echo -e "DESCRIPTION: This script verifies the current SDP setup for the specified instance, and also performs basic health checks of configured servers. This uses the SDP instance bin directory /p4/N/bin to determine what server binaries (p4d, p4broker, p4p) are expected to be configured on this machine. Existence of the '*_init' script indicates the given binary is expected. For example, for instance 1, if /p4/1/bin/p4d_1_init exists, a p4d server is expected to run on this machine. Checks may be executed or skipped depending on what servers are configured. For example, if a p4d is configured, the \$P4ROOT/server.id file should exist. If p4p is configured, the 'cache' directory should exist. OPTIONS: <instance> Specify the SDP instances. If not specified, the SDP_INSTANCE environment variable is used instead. If the instance is not defined by a parameter and SDP_INSTANCE is not defined, exits immediately with an error message. -online Online mode. Does additional checks that require P4D to be online. -skip <test>[,<test2>,...] Specify a comma-delimited list of test names to skip. Valid test names: * crontab: Skip crontab check. Use this if you do not expect crontab to be configured, perhaps if you use a different scheduler. * excess: Skip checks for excess copies of p4d/p4p/p4broker in PATH. * license: Skip license related checks. * offline_db: Skip checks that require a healthy offline_db. * masterid: Skip check to ensure ServerID of master starts with 'master'. * version: Skip version checks. As an alternative to using the '-skip' command, the shell environment variable VERIFY_SDP_SKIP_TEST_LIST can be set to a comma-separated list of test names to skip. Using the command line parameter is the best choice for temporarily skipping tests, while specifying the environment variable is better for making permanent exceptions (e.g. always excluding the crontab check if crontabs are not used at this site). The variable should be set in /p4/common/config/p4_N.vars. If the '-skip' option is provided, the VERIFY_SDP_SKIP_TEST_LIST variable is ignored (not appended to). So it may make sense to reference the variable on the command line. For example, if the value of the variable is 'crontab', to skip crontab and license checks, you could specify: -skip \$VERIFY_SDP_SKIP_TEST_LIST,license -si Silent mode, useful for cron operation. Both stdout and stderr are still captured in the log. The '-si' option cannot be used with '-L off'. -L <log> Specify the log file to use. The default is /p4/N/logs/verify_sdp.log The special value 'off' disables logging to a file. Note that '-L off' and '-si' are mutually exclusive. -D Set extreme debugging verbosity. HELP OPTIONS: -h Display short help message -man Display man-style help message EXAMPLES: Example 1: Typical usage: This script is typically called after SDP update with only the instance name or number as an argument, e.g.: verify_sdp.sh 1 Example 2: Skipping some checks. verify_sdp.sh 1 -skip crontab Example 3: Automation Usage If used from automation already doing its own logging, use -L off: verify_sdp.sh 1 -L off LOGGING: This script generates a log file and also displays it to stdout at the end of processing. By default, the log is: /p4/N/logs/verify_sdp.log. The exception is usage errors, which result an error being sent to stderr followed usage info on stdout, followed by an immediate exit. If the '-si' (silent) flag is used, the log is generated, but its contents are not displayed to stdout at the end of processing. EXIT CODES: An exit code of 0 indicates no errors were encounted attempting to perform verifications, and that all checks verified cleanly. " # end::includeManual[] fi exit 1 } #------------------------------------------------------------------------------ # Function: terminate function terminate { # Disable signal trapping. trap - EXIT SIGINT SIGTERM [[ "$Log" == "off" ]] || msg "\\nLog is: $Log\\n${H1}\\n" # With the trap removed, exit. exit "$ErrorCount" } #------------------------------------------------------------------------------ # Function: do_preflight_checks ($instance) # # If preflight checks fail, further tests are aborted. Failure of the very # basic preflight checks is an indication that the SDP structure is in # need of repair. # # Sample Usage: # do_preflght_checks "$SDP_INSTANCE" ||\ # bail "Preflight checks failed. Aborting further checks." #------------------------------------------------------------------------------ function do_preflight_checks () { local instance="${1:-}" local toolsList="awk date grep head id ls tail tee which" msg "$H2\\nDoing preflight sanity checks." msg "Preflight Check: Ensuring these utils are in PATH: $toolsList" for tool in $toolsList; do CheckCount+=1 [[ -z "$(command -v "$tool")" ]] && \ errmsg "Tool '$tool' not in PATH." done [[ $ErrorCount -eq 0 ]] || return 1 msg "Verified: Essential tools are in the PATH." msg "Preflight Check: cd $VS_SDP_P4CBIN" CheckCount+=1 if cd "$VS_SDP_P4CBIN"; then cd "$OLDPWD" || bail "Failed to cd to $OLDPWD. Aborting." else errmsg "Could not cd to: $VS_SDP_P4CBIN" return 1 fi msg "Verified: cd works to: $VS_SDP_P4CBIN" msg "Preflight Check: Checking current user owns $VS_SDP_P4CBIN" # shellcheck disable=SC2012 VS_SDP_OWNER="$(ls -ld . | awk '{print $3}')" ThisUser="$(id -n -u)" CheckCount+=1 if [[ "$ThisUser" == "$VS_SDP_OWNER" ]]; then msg "Verified: Current user [$ThisUser] owns $VS_SDP_P4CBIN" else errmsg "Current user [$ThisUser] does not own $VS_SDP_P4CBIN. This most likely means this script is running as the wrong user. It could also mean the $VS_SDP_P4CBIN directory is not owned by by the correct owner, which should be the OS account under which the p4d process runs." return 1 fi msg "Preflight Check: Checking /p4 and /p4/<instance> are local dirs." if ! check_local_instance_home_dir "$instance"; then errmsg "The SDP /p4 and /p4/<instance> dirs are NOT local." return 1 fi return 0 } #------------------------------------------------------------------------------ # Function: check_file ($file, $errMsg, $warningOnly) # # Checks for existence of a file. Returns 0 if it exists, 1 otherwise. # # Allows optional custom error message describing the file, to be displayed if # the file is missing. Default error message is "Missing file [FILE]." #------------------------------------------------------------------------------ function check_file () { local file=$1 local errMsg=${2:-Missing file} local warningOnly=${3:-0} CheckCount+=1 msg "Checking existence of file $file" if [[ "$warningOnly" -eq 0 ]]; then [[ -f "$file" ]] && return 0 errmsg "$errMsg: [$file]." else [[ -f "$file" ]] && return 0 warnmsg "$errMsg: [$file]." fi return 1 } #------------------------------------------------------------------------------ # Function: check_file_x ($file, $errMsg, $warningOnly) # # Checks existence of a file with executable bit set. Returns 0 if it exists # and is executable, 1 otherwise. # # Allows optional custom error message describing the file, to be displayed if # the file is missing. Default error message is "File not executable [FILE]." #------------------------------------------------------------------------------ function check_file_x () { local file=$1 local errMsg=${2:-File not executable} local warningOnly=${3:-0} CheckCount+=1 msg "Checking executable file $file" if [[ "$warningOnly" -eq 0 ]]; then [[ -x "$file" ]] && return 0 errmsg "$errMsg: [$file]." else [[ -x "$file" ]] && return 0 warnmsg "$errMsg: [$file]." fi return 1 } #------------------------------------------------------------------------------ # Function: check_file_dne ($file, $errMsg, $warningOnly) # # Confirm that a specific file does not exist, e.g. a semaphore file. If the # specified files does not exist, return 0, or 1 if it exists. This is the # opposite of check_file(). # # Allows optional custom error message describing the file, to be displayed if # the file is found. Default error message is "This file should not exist: [FILE]." #------------------------------------------------------------------------------ function check_file_dne () { local file=$1 local errMsg=${2:-This file should not exist} local warningOnly=${3:-0} CheckCount+=1 msg "Confirming this file does not exist: $file" if [[ "$warningOnly" -eq 0 ]]; then [[ ! -f "$file" ]] && return 0 errmsg "$errMsg: [$file]." else [[ ! -f "$file" ]] && return 0 warnmsg "$errMsg: [$file]." fi return 1 } #------------------------------------------------------------------------------ # Function: check_configurable ($instance, $configurable, $scope, $expectedVal, $errMsg1, $errMsg2) # # Check that a configurable is set, and optionally check that it is set to # an expected value. # # Inputs: # $1 - SDP Instance. Required. # $2 - Configurable name. Required. # $3 - Configurable scope/ServerID, as per 'p4 help configure'. The default # is "any", meaning what it means with 'p4 configure set', i.e. that the # configurable is a global default. The special value 'ALL' can # also be supplied parameter, which is has the special meaning of checking # if the configurable is defined for any ServerID, including the 'any' # value. The value returned is that of the first setting encountered. # $4 - Expected value of configurable. Optional. If defined, an additional check is # done, checking the current value against the expected value. Optionally, # the special value UNDEF can be used, which reverses the exit code, such # that a happy zero is returned only if the value is not set. # $5 - Optional error message to display if no value is defined. See code # below for the default message. # $6 - Optional error message to display if a value is defined but does not # match the expected value. See code below for the default message. # # Return Codes: # 1 - Verifications failed. # 0 - Verifications passed. # # Sample Usage: # check_configurable "$SDP_INSTANCE" journalPrefix # # check_configurable "$SDP_INSTANCE" journalPrefix any "$CHECKPOINTS/$P4SERVER" # # check_configurable "$SDP_INSTANCE" journalPrefix any "$CHECKPOINTS/$P4SERVER" ||\ # bail "Yikes, journalPrefix is not set, all bets are off. Aborting." #------------------------------------------------------------------------------ function check_configurable () { local instance="$1" local configurable="$2" local scope="${3:-any}" local expectedValue="${4:-NoExpectedValue}" local errMsgMissing="${5:-No value defined}" local errMsgBadValue="${6:-Value does not match what is expected}" local detectedScope= local value= CheckCount+=1 if [[ ! -r "$P4ROOT"/db.config ]]; then warnmsg "Skipping check for configurable $configurable; no db.config." return 1 fi if [[ "$scope" != "ALL" ]]; then value=$($P4DBIN -r "$P4ROOT" -cshow | grep "^${scope}: ${configurable} = ") else value=$($P4DBIN -r "$P4ROOT" -cshow | grep ": ${configurable} = " | head -1) detectedScope="$value" value=${value##* = } detectedScope="${detectedScope%%:*}" fi if [[ "$expectedValue" != "UNDEF" ]]; then if [[ -n "$value" ]]; then value=${value##* = } if [[ "$scope" != "ALL" ]]; then msg "Verified: Configurable ${scope}:${configurable} is defined." else msg "Verified: Configurable ${configurable} is defined for at least once." fi else errmsg "$errMsgMissing for configurable [${scope}:${configurable}]." return 1 fi else if [[ -n "$value" ]]; then if [[ "$scope" != "ALL" ]]; then errmsg "Configurable ${configurable} should not be set with 'p4 configure set' but has a value for ServerID ${scope} of: ${value}" return 1 else errmsg "Configurable ${configurable} should not be set with 'p4 configure set' but has a value for ServerID ${detectedScope} of: ${value} (and possibly for other ServerIDs)." return 1 fi else if [[ "$scope" != "ALL" ]]; then msg "Verified: Configurable ${scope}:${configurable} is undefined." else msg "Verified: Configurable ${configurable} is undefined." fi fi fi [[ "$expectedValue" == "NoExpectedValue" ]] && return 0 CheckCount+=1 if [[ "$expectedValue" != "UNDEF" ]]; then if [[ "$value" == "$expectedValue" ]]; then msg "Verified: Configurable ${scope}:${configurable} has expected value [$value]." else errmsg "$errMsgBadValue for variable [${scope}:${configurable}]\\n\\tExpected value: [$expectedValue]\\n\\tActual value: [$value]" return 1 fi fi return 0 } #------------------------------------------------------------------------------ # Function: check_env_var ($instance, $var, $expectedval, $msg1, $msg2) # # Check that a shell environment variable is set when sourcing the SDP # environment. Optionally checks that variables are set to expected values. # # Inputs: # $1 - SDP Instance. Required. # $2 - Variable name. Required. # $3 - Expected value of variable. Optional. If defined, an additional check is # done, checking the current value against the expected value. # $4 - Optional error message to display if no value is defined. See code # below for the default message. # $5 - Optional error message to display if a value is defined but does not match # the expected value. See code below for the default message. # # Return Codes: # 1 - Verifications failed. # 0 - Verifications passed. # Sample Usage: # check_env_var $SDP_INSTANCE P4JOURNAL "/p4/$SDP_INSTANCE/logs/journal" # # check_env_var $SDP_INSTANCE P4JOURNAL "/p4/$SDP_INSTANCE/logs/journal" ||\ # bail "Yikes, P4JOURNAL is not set, all bets are off. Aborting." #------------------------------------------------------------------------------ function check_env_var () { local instance="$1" local var="$2" local expectedValue="${3:-NoExpectedValue}" local errMsgMissing="${4:-No value defined}" local errMsgBadValue="${5:-Value does not match what is expected}" local value= CheckCount+=1 eval unset "${var}" # shellcheck disable=SC1090 source "$VS_SDP_ENV" "$instance" set +u if [[ -n "$(eval echo \$"${var}")" ]]; then msg "Verified: Variable ${var} is defined." set -u else errmsg "$errMsgMissing for variable [$var]." set -u return 1 fi [[ "$expectedValue" == "NoExpectedValue" ]] && return 0 CheckCount+=1 value="$(eval echo \$"${var}")" if [[ "$value" == "$expectedValue" ]]; then msg "Verified: Variable ${var} has expected value [$value]." else errmsg "$errMsgBadValue for variable [$var]\\n\\tExpected value: [$expectedValue]\\n\\tActual value: [$value]" return 1 fi return 0 } #------------------------------------------------------------------------------ # Function: check_local_instance_home_dir ($instance) # # Check that the '/p4' directory and the instance home directory '/p4/N' are # local directories on the root volume, per SDP structural intent. # # Inputs: # $1 - SDP Instance. Required. # # Return Codes: # 0 - Verifications were able to at least run; ErrorCount is incremented # if tests fail. # 1 - Verifications could not even complete. This is a pre-flight failure. # # This increments globals CheckCount and possibly ErrorCount. # # Sample Usage: # check_local_instance_home_dir "$SDP_INSTANCE" ||\ # bail "Error checking p4dir and/dor local instance home dir." #------------------------------------------------------------------------------ function check_local_instance_home_dir () { local instance="$1" local p4Dir="/p4" local p4HomeDir="$p4Dir/$instance" CheckCount+=1 if [[ "$P4HOME" == "$p4HomeDir" ]]; then msg "Verified: P4HOME has expected value: $p4HomeDir" else errmsg "P4HOME has unexpected value: $p4HomeDir" fi CheckCount+=1 if [[ -L "$p4HomeDir" ]]; then errmsg "This is a symlink; it should be a local directory: $p4HomeDir" else msg "Verified: This P4HOME path is not a symlink: $p4HomeDir" fi CheckCount+=1 if cd "$p4Dir"; then msg "Verified: cd to $p4Dir OK." CheckCount+=1 if [[ "$(pwd -P)" == "$p4Dir" ]]; then msg "Verified: Dir $p4Dir is a local dir." else errmsg "Dir $p4Dir is NOT a local dir." fi cd - > /dev/null || bail "Failed to cd to $OLDPWD. Aborting." CheckCount+=1 if cd "$p4HomeDir"; then msg "Verified: cd to $p4HomeDir OK." CheckCount+=1 if [[ "$(pwd -P)" == "$p4HomeDir" ]]; then msg "Verified: P4HOME dir $p4HomeDir is a local dir." else errmsg "P4HOME dir $p4HomeDir is NOT a local dir." fi cd - > /dev/null || bail "Failed to cd to $OLDPWD. Aborting." else errmsg "Could not cd to $p4HomeDir." return 1 fi else errmsg "Could not cd to $p4Dir." return 1 fi return 0 } #============================================================================== # Command Line Processing declare -i shiftArgs=0 set +u while [[ $# -gt 0 ]]; do case $1 in (-h) usage -h;; (-man) usage -man;; (-online) ServerOnline=1;; (-skip) SkipTestList="$2"; shiftArgs=1;; (-si) SilentMode=1;; (-L) Log="$2"; shiftArgs=1;; (-D) set -x;; # Debug; use 'set -x' mode. (-*) usage -h "Unknown command line option ($1).";; (*) export SDP_INSTANCE=$1;; esac # Shift (modify $#) the appropriate number of times. shift; while [[ $shiftArgs -gt 0 ]]; do [[ $# -eq 0 ]] && usage -h "Incorrect number of arguments." shiftArgs=$shiftArgs-1 shift done done set -u #============================================================================== # Command Line Verification [[ "$SDP_INSTANCE" == "UnsetSDPInstance" ]] && \ usage -h "Missing <instance> parameter. The <instance> must be given as a parameter to this script, or else the \$SDP_INSTANCE environment variable defined. It can be set by doing:\\n\\n\\tsource $VS_SDP_ENV <instance>\\n\\nor by passing in the instance name as a parameter to this script.\\n" [[ "$SilentMode" -eq 1 && "$Log" == "off" ]] && \ usage -h "Cannot use '-si' with '-L off'." #============================================================================== # Main Program # shellcheck disable=SC1090 source "$VS_SDP_ENV" "$SDP_INSTANCE" ||\ bail "Failed to load SDP environment for instance $SDP_INSTANCE." # shellcheck disable=SC1090 source "$P4CBIN/backup_functions.sh" ||\ bail "Failed to load backup_functions.sh." [[ "${OSUSER:-Unset}" == "Unset" ]] &&\ bail "The critical OSUSER setting is not defined in $VS_SDP_ENV. Aborting." # If this verify_sdp.sh script is called by root, change user to OSUSER. if [[ $(id -u) -eq 0 ]]; then exec su - "$OSUSER" -c "$VS_SDP_P4CBIN/${0##*/} $CmdArgs" elif [[ $(id -u -n) != "${OSUSER:-UnknownOSUSER}" ]]; then bail "${0##*/} can only be run by root or $OSUSER" fi trap terminate EXIT SIGINT SIGTERM if [[ -z "$SkipTestList" && -n "${VERIFY_SDP_SKIP_TEST_LIST:-}" ]]; then SkipTestList="$VERIFY_SDP_SKIP_TEST_LIST" fi if [[ -n "$SkipTestList" ]]; then for test in $(echo "$SkipTestList" | tr ',' ' '); do case "$test" in (cron|crontab) DoCrontabTest=0;; (excess) DoExcessBinaryTest=0;; (license) DoLicenseTest=0;; (masterid) DoMasterIDTest=0;; (offline_db) DoOfflineDBTest=0;; (version) DoVersionTest=0;; (*) errmsg "Invalid test name specified with '-skip': $test";; esac done fi # Logs should be defined to /p4/N/logs after sourcing the environment # file above; default to /tmp for cases of incomplete environment where # LOGS is not defined. export LOGS="${LOGS:-/tmp}" [[ "$Log" == "Unset" ]] && Log="${LOGS}/verify_sdp.log" if [[ "$Log" != "off" ]]; then if [[ -f "$Log" ]]; then if [[ ! -w "$Log" ]]; then BadLog="$Log" Log="off" bail "Existing log file [$BadLog] is not writable. Aborting." fi rotate_log_file "$Log" ".gz" else if [[ ! -d "${LOGS}" ]]; then Log="off" bail "Logs directory [$LOGS] is not writable. Aborting." fi fi if ! touch "$Log"; then BadLog="$Log" Log="off" bail "Couldn't touch log file [$BadLog]. Aborting." fi # Redirect stdout and stderr to a log file. if [[ "$SilentMode" -eq 0 ]]; then exec > >(tee "$Log") exec 2>&1 else exec >"$Log" exec 2>&1 fi msg "${H1}\\nLog is: $Log" fi [[ "$P4TMP" != "Unset" && -d "$P4TMP" && -w "$P4TMP" ]] ||\ bail "SDP environment must define required P4TMP variable. Value must be a directory that is writable; value is: $P4TMP" msg "$ThisScript v$Version Starting SDP verification on host ${HOSTNAME%%.*} at $(date +'%a %Y-%m-%d %H:%M:%S %Z') with this command line:\\n$CmdLine" msg "\\nIf you have any questions about the output from this script, contact support@perforce.com." do_preflight_checks "$SDP_INSTANCE" ||\ bail "Preflight checks failed. Aborting further checks." msg "${H2}\\nChecking environment variables." check_env_var "$SDP_INSTANCE" SDP_INSTANCE check_env_var "$SDP_INSTANCE" P4ROOT "/p4/$SDP_INSTANCE/root" check_env_var "$SDP_INSTANCE" P4JOURNAL "/p4/$SDP_INSTANCE/logs/journal" check_env_var "$SDP_INSTANCE" P4MASTER_ID check_env_var "$SDP_INSTANCE" P4MASTERHOST check_env_var "$SDP_INSTANCE" P4MASTERPORT check_env_var "$SDP_INSTANCE" SERVERID if ! check_env_var "$SDP_INSTANCE" SDP_ADMIN_PASSWORD_FILE "$P4CCFG/.p4passwd.$P4SERVER.admin"; then check_file "$SDP_ADMIN_PASSWORD_FILE" "SDP admin password file doesn't exist" fi msg "${H2}\\nRunning standard checks typically called within SDP scripts." check_vars CheckCount+=1 set_vars if [[ "$P4DServer" -eq 1 ]]; then check_dirs CheckCount+=1 fi msg "${H2}\\nChecking *_init scripts in instance bin dir [/p4/$SDP_INSTANCE/bin] to see what servers are configured on this machine." # shellcheck disable=SC2154 if [[ -e "$P4DInitScript" ]]; then msg "A p4d server is here." P4DServer=1 check_file_x "$P4DInitScript" fi # shellcheck disable=SC2154 if [[ -e "$P4BrokerInitScript" ]]; then msg "A p4broker server is here." P4BrokerServer=1 check_file_x "$P4BrokerInitScript" fi # shellcheck disable=SC2154 if [[ -e "$P4ProxyInitScript" ]]; then msg "A p4p server is here." P4ProxyServer=1 check_file_x "$P4ProxyInitScript" fi if [[ $((P4DServer+P4BrokerServer+P4ProxyServer)) -eq 0 ]]; then CheckCount+=1 errmsg "No servers (p4d, p4p, p4broker) are configured." fi if [[ -d /etc/system/system ]]; then msg "${H2}\\nChecking that systemd service files are readable." # shellcheck disable=SC2154 if [[ -e "$P4DSystemdServiceFile" ]]; then CheckCount+=1 if systemctl cat "${P4DBIN##*/}" > /dev/null 2>&1; then msg "Verified: Permissions OK to status w/o sudo for p4d service." else errmsg "Could not run 'systemctl status ${P4DBIN##*/}'." fi fi # shellcheck disable=SC2154 if [[ -e "$P4BrokerSystemdServiceFile" ]]; then CheckCount+=1 if systemctl cat "${P4BROKERBIN##*/}" > /dev/null 2>&1; then msg "Verified: Permissions OK to status w/o sudo for p4broker service." else errmsg "Could not run 'systemctl status ${P4BROKERBIN##*/}'." fi fi # shellcheck disable=SC2154 if [[ -e "$P4ProxySystemdServiceFile" ]]; then CheckCount+=1 if systemctl cat "${P4PBIN##*/}" > /dev/null 2>&1; then msg "Verified: Permissions OK to status w/o sudo for p4p service." else errmsg "Could not run 'systemctl status ${P4PBIN##*/}'." fi fi fi if [[ "$P4DServer" -eq 1 ]]; then msg "${H2}\\nChecking for a few database files." # Check db files exist for file in db.counters db.domain db.config; do check_file "$P4ROOT/$file" "Expected database file doesn't exist." done if [[ "$DoOfflineDBTest" -eq 1 ]]; then for file in db.counters db.domain db.config; do check_file "$OFFLINE_DB/$file" "Expected database file doesn't exist." done fi fi msg "${H2}\\nChecking for existence of key files." check_file "$P4BIN" "The p4 binary (or symlink) doesn't exist." check_file "$P4TICKETS" "The P4TICKETS file doesn't exist." if [[ "$P4DServer" -eq 1 ]]; then check_file "$P4ROOT/server.id" "The required $P4ROOT/server.id file is missing." check_file_dne "$P4ROOT/P4ROOT_not_usable.txt" "P4ROOT is not in a usable state." if [[ "$DoOfflineDBTest" -eq 1 ]]; then check_file "$OFFLINE_DB/offline_db_usable.txt" "Offline database not in a usable state." check_file_dne "$OFFLINE_DB/P4ROOT_not_usable.txt" "Offline database has P4ROOT_not_usable.txt file." fi if [[ "$DoLicenseTest" -eq 1 ]]; then msg "${H2}\\nLicense Checks." #-------------------------------------------------------------------------- # A sample license line looks like this: # License: Perforce Battle School 28 users (support ends 2020/06/01) (expires 2020/06/01) # Existence of 'support ends' indicates the license is valid. # Existence of 'expires' indicates a temp or subscription license. #-------------------------------------------------------------------------- if check_file "$P4ROOT/license" "The \$P4ROOT/license file doesn't exist" 1; then LicenseInfo=$($P4DBIN -V | grep '^License:') CheckCount+=1 if [[ -n "$LicenseInfo" ]]; then if [[ "$LicenseInfo" == *" expired "* ]]; then errmsg "The license is expired." elif [[ "$LicenseInfo" == *"expires"* ]]; then LicenseExpiration=${LicenseInfo##*(expires } LicenseExpiration=${LicenseExpiration%%)*} # Check License CheckCount+=1 CurrentTime=$(date +%s 2>/dev/null) ExpirationTime=$(date +%s --date "$LicenseExpiration" 2>/dev/null) if [[ -n "$CurrentTime" && -n "$ExpirationTime" ]]; then TimeDiff=$((ExpirationTime-CurrentTime)) DaysDiff=$((TimeDiff/(3600*24))) msg "Info: License expires on $LicenseExpiration (in $DaysDiff days)." if [[ "$DaysDiff" -le "$LicenseDaysExpireWarning" ]]; then errmsg "License will expire within $LicenseDaysExpireWarning days." fi else msg "Skipping license check due to incompatible 'date' utility on this OS." fi elif [[ "$LicenseInfo" == *"support ends"* ]]; then msg "Info: License is perpetual." fi else errmsg "Could not determine license info from license file." fi fi else msg "Skipping license check per '-skip'." fi fi if [[ "$P4DServer" -eq 1 ]]; then msg "${H2}\\nChecking configurables values." check_configurable "$SDP_INSTANCE" P4JOURNAL ALL UNDEF msg "Note: P4JOURNAL must be defined as a shell environment variable and must not set in db.config.\\n" check_configurable "$SDP_INSTANCE" journalPrefix any "$CHECKPOINTS/$P4SERVER" check_configurable "$SDP_INSTANCE" server.depot.root any "$DEPOTS" fi if [[ "$P4DServer" -eq 1 ]]; then msg "${H2}\\nChecking P4ROOT and offline_db syminks." CheckCount+=1 LinkP4ROOT="$(readlink "$P4ROOT")" if [[ "$LinkP4ROOT" == *"/db1" || "$LinkP4ROOT" == *"/db2" ]]; then msg "Verified: Symlink for P4ROOT points to a db1 or db2 path." else errmsg "Symlink for P4ROOT does not point to a db1 or db2 path." fi if [[ "$DoOfflineDBTest" -eq 1 ]]; then CheckCount+=1 LinkOfflineDB="$(readlink "$OFFLINE_DB")" if [[ "$LinkOfflineDB" == *"/db1" || "$LinkOfflineDB" == *"/db2" ]]; then msg "Verified: Symlink for offline_db points to a db1 or db2 path." else errmsg "Symlink for offline_db does not point to a db1 or db2 path." fi fi CheckCount+=1 if [[ "$LinkP4ROOT" != "$LinkOfflineDB" ]]; then msg "Verified: Symlinks for P4ROOT and offline_db do not point to the same target." else errmsg "Symlinks for P4ROOT and offline_db point to the same target." fi fi msg "${H2}\\nChecking for standard symlink/dir structure." CheckCount+=1 if cd "$P4HOME"; then CheckCount+=1 if [[ -d "$PWD/bin" ]]; then msg "Verified: $PWD/bin is a regular directory." else warnmsg "$PWD/bin is not a regular directory." fi linkList="logs tmp" [[ "$P4DServer" -eq 1 ]] && linkList+=" checkpoints depots root offline_db" [[ "$P4ProxyServer" -eq 1 ]] && linkList+=" cache" for link in $linkList; do CheckCount+=1 if [[ -L "$PWD/$link" ]]; then msg "Verified: $PWD/$link is a symlink." else errmsg "$PWD/$link is not a symlink directory." fi done CheckCount+=1 if [[ -L "/p4/common" ]]; then msg "Verified: /p4/common is a symlink." else errmsg "Path /p4/common is not a symlink." fi CheckCount+=1 if [[ -L "/p4/sdp" ]]; then msg "Verified: /p4/sdp is a symlink." else errmsg "Path /p4/sdp is not a symlink." fi cd - > /dev/null || bail "Failed to cd to $OLDPWD. Aborting." else errmsg "Could not cd to $P4HOME." fi if [[ "$ServerOnline" -eq 1 ]]; then msg "${H2}\\nDoing online checks." CheckCount+=1 if run "$P4BIN -s info -s" "Checking 'p4 -s info -s'" 1; then msg "Verified: 'p4 -s info -s' output is nominal." else errmsg "Could not verify the p4d server is online (P4PORT=$P4PORT)." fi CheckCount+=1 if run "$P4BIN -s -p $P4MASTERPORT info -s" "Checking 'p4 -s info -s'" 1; then msg "Verified: 'p4 -s -p $P4MASTERPORT info -s' output is nominal." else errmsg "Could not verify the p4d server is online (P4MASTERPORT=$P4MASTERPORT)." fi CheckCount+=1 if run "$P4CBIN/p4login" "Login check." 0; then msg "Verified: Login OK." else errmsg "Login as P4USER $P4USER to P4PORT $P4PORT could not be verified." fi CheckCount+=1 if run "$P4CBIN/p4login -p $P4MASTERPORT" "Login check using P4MASTERPORT." 0; then msg "Verified: Login OK using P4MASTERPORT." else errmsg "Login as P4USER $P4USER to P4MASTERPORT $P4MASTERPORT could not be verified." fi fi if [[ "$DoCrontabTest" -eq 1 ]]; then msg "${H2}\\nChecking crontab for user $USER." CheckCount+=1 TmpFile="$(mktemp "$P4TMP/crontab.XXXXXXXXXXX")" if crontab -l | grep -v '^#' > "$TmpFile"; then if grep -q -E '/(daily_checkpoint.sh|replica_status.sh) ' "$TmpFile"; then msg "Verified: Crontab for user $USER passed basic sanity check." else errmsg "Crontab for user $USER did not pass basic sanity check; missing call to daily_checkpoint.sh or replica_status.sh" fi else errmsg "Could not capture crontab." fi rm -f "$TmpFile" else msg "Skipping crontab check per '-skip'." fi if [[ "$DoVersionTest" -eq 1 ]]; then msg "${H2}\\nChecking SDP Version Id (current and legacy methods)." CheckCount+=1 if [[ -r /p4/sdp/Version ]]; then SDPVersionA=$(get_sdp_version_from_string "$(cat /p4/sdp/Version)") else SDPVersionA="Unknown" errmsg "Missing SDP Version file: /p4/sdp/Version" fi CheckCount+=1 SDPVersionB=$(get_sdp_version_from_string "$SDP_VERSION") if [[ "$SDPVersionA" == "$SDPVersionB" ]]; then msg "SDP Version from /p4/common/bin/p4_vars matches." else errmsg "SDP Version from /p4/common/bin/p4_vars isn't set or doesn't match /p4/sdp/Version:\\n[$SDPVersionB] vs. [$SDPVersionA]" fi if [[ "$ServerOnline" -eq 1 ]]; then CheckCount+=1 SDPVersionC=$(get_sdp_version_from_string "$("$P4BIN" counter SDP_VERSION)") if [[ "$SDPVersionA" == "$SDPVersionC" ]]; then msg "SDP Version from 'p4 counter SDP_VERSION' matches." else if [[ "$SDPVersionC" == "Unknown" ]]; then warnmsg "SDP Version from 'p4 counter SDP' is not set." else errmsg "SDP Version from 'p4 counter SDP' doesn't match /p4/sdp/Version:\\n[$SDPVersionC] vs. [$SDPVersionA]" fi fi fi else msg "Skipping version checks per '-skip'." fi if [[ "$DoExcessBinaryTest" -eq 1 ]]; then msg "${H2}\\nChecking for Helix executables outside /p4/common/bin in PATH." CheckCount+=1 # Check for excess Helix server binaries in PATH, and classify that as an error. for exe in p4d p4broker p4p; do for exeInPath in $(which -a $exe); do if [[ "${exeInPath%/*}" != *"$P4CBIN" ]]; then errmsg "Executable $exe found outside /p4/common/bin in PATH: ${exeInPath%/*}" ExcessServerBinariesFound=1 fi done done # Check for excess Helix client binary in PATH, and classify that as a warning. for exeInPath in $(which -a p4); do if [[ "${exeInPath%/*}" != *"$P4CBIN" ]]; then warnmsg "Executable 'p4' found outside /p4/common/bin in PATH: ${exeInPath%/*}" fi done if [[ "$ExcessServerBinariesFound" -eq 0 ]]; then msg "Verified: No excess Helix server binaries found outside /p4/common/bin." fi else msg "Skipping excess binary checks per '-skip'." fi if [[ "$DoMasterIDTest" -eq 1 ]]; then msg "${H2}\\nChecking that P4MASTER_ID value starts with 'master'." CheckCount+=1 if [[ "$P4MASTER_ID" == "master"* ]]; then msg "Verified: The P4MASTER_ID value starts with 'master'." else errmsg "The P4MASTER_ID value ($P4MASTER_ID) does not start with 'master'." fi fi # TODO: # - Ensure checkpoints dir contains a checkpoint or two. # - service password verified as set for a replica (value doesn't matter, it just needs to be set). # - Add flag to check less-critical SDP configurables, and generate # warnings (rather than errors) if they are not set as expected, # using SDP configure_new_servers.sh script as a guide. if [[ "$ErrorCount" -eq 0 && "$WarningCount" -eq 0 ]]; then msg "\\n${H1}\\n\\nALL CLEAN: $CheckCount verifications completed OK." elif [[ "$ErrorCount" -eq 0 ]]; then msg "\\n${H1}\\n\\nNO ERRORS: $CheckCount verifications completed, with $WarningCount warnings detected." else msg "\\n${H1}\\n\\nVerifications completed, with $ErrorCount errors and $WarningCount warnings detected in $CheckCount checks." fi # See the terminate() function, which is really where this script exits. exit 0
# | Change | User | Description | Committed | |
---|---|---|---|---|---|
#58 | 30933 | C. Thomas Tyler |
Added check for proper ownership of all files in P4ROOT. Added separate utility for checking ownership (and optionally group) of all files in a given path. Fixes SDP-1119. #review-30934 |
||
#57 | 30793 | C. Thomas Tyler | Expanded crontab check to allow for a standalone broker install. | ||
#56 | 30270 | Robert Cowham | Fix shellcheck warnings and use of copy_jd_table | ||
#55 | 30267 | Robert Cowham |
Copy files to be dumped via p4d -jd to tmp dir first to avoid locks on P4ROOT (or offline_db) SDP-1087 |
||
#54 | 30131 | C. Thomas Tyler |
In verify_sdp.sh: * The P4MASTER_ID check now accepts 'commit' or 'master'. * The 'masterid' skippable check is now synonymous with 'commitid' #review-30132 |
||
#53 | 30022 | C. Thomas Tyler | Minor doc clarification in verify_sdp.sh | ||
#52 | 30021 | C. Thomas Tyler |
Enhanced docs for ccheck.sh script. Added examples. Added '-c' option to verify_sdp.sh to do configurables check. |
||
#51 | 29121 | C. Thomas Tyler |
Fixed dependency on cwd if dot is in PATH in check for exes. #review-29003 |
||
#50 | 29100 | C. Thomas Tyler |
Removed requirement for systemd *.service files to have open perms in *_base scripts and templates. Removed checks for open perms on *.service files in verify_sdp.sh. Fixed minor ShellCheck compliance issue. See also: HI-101: https://swarm.workshop.perforce.com/jobs/HI-101 [Submitting, then re-opening for post-commit final review]. |
||
#49 | 28815 | C. Thomas Tyler |
Enhanced error message in verify_sdp.sh if p4d_N script is malformed. The /p4/N/bin/p4d_N only exists as a script for case-insensitive servers. Normally this file is generated (by mkdirs.sh), but in some DIY installations it may be handcrafted, and possibly wrong. This changes improves the error message in that scenario, such that the user has enough information to correct the problem. This also fixes a funky formatting issue when the error occurred. #review-28816 |
||
#48 | 28771 | C. Thomas Tyler |
Changed email address for Perforce Support. #review-28772 @amo @robert_cowham |
||
#47 | 28641 | C. Thomas Tyler |
Tweaked scripts to support IPv6 SSL prefixes. Added test script to test bash code snippets. First test is the snippet to check if SSL is enabled, and if so get the SSL prefix. |
||
#46 | 28553 | C. Thomas Tyler |
Fixed issue file type detection (shell script vs. binary) did not work on some Linux platforms, due to slightly different output from 'file' on shell scripts on different platforms. Type detection logic is now more robust. The test suite will catch this when we add support for Rocky Linux 8. Also corrected a copy/paste error in an error message and enhanced code comments. This is a bug fix to a new, not-yet-released feature. #review-28554 |
||
#45 | 28423 | C. Thomas Tyler |
verify_sdp.sh v5.21.0: Added check that simply named files p4, p4d, p4p, and p4broker in /p4/common/bin are indeed shell scripts, not binaries. Also silenced configurable checks and some other checks that should be skipped if '-skip p4root' is specified. #review-28424 @russell_jackson |
||
#44 | 28421 | C. Thomas Tyler |
verify_sdp.sh v5.20.0: * New checks: /p4/N/bin/p4{d,p,broker}_N need correct symlink target, and must exist if the corresponding _init script exists. For p4d, it can be a symlink (for a case-sensitive instance) or script (for a case-insensitive instance to pass the C1 flag). Either way the target is checked. These checks cannot be skipped or converted to warnings. * Added check that /p4/N/bin/p4{d,p,broker}_N_init scripts have content that matches templates. This can be skipped with '-skip' or reported as mere warnings (with '-warn') with a new and documented 'init' category of test skipping/warning. #review-28422 |
||
#43 | 28250 | C. Thomas Tyler | SDP Version checks for 'p4 counter SDP_VERSION' downgraded to warning. | ||
#42 | 28134 | C. Thomas Tyler |
Added '-warn' option to verify_sdp.sh to convert errors to warnings. This option takes a list of named tests, using the same set of tests names as used with the existing '-skip' option. Rather than skipping the tests entirely, with '-warn' wany would-be errors are instead reported as warnings. Warnings will not cause a non-zero exit status. A new environment variable is added to go with the new option. Similar to how the existing '-skip' option has a corresponding VERIFY_SDP_SKIP_TEST_LIST setting the Instance Vars file, the -warn option has a corresponding $VERIFY_SDP_WARN_TEST_LIST. |
||
#41 | 28047 | C. Thomas Tyler |
Corrected counter name re: SDP_VERSION counter, and added supplemental information on how to fix it. |
||
#40 | 27961 | C. Thomas Tyler | Corrected typo in error message in verify_sdp.sh. | ||
#39 | 27918 | C. Thomas Tyler |
Corrected typo in a comment, and minor code style tweaks. This is a non-functional change to verify_sdp.sh. |
||
#38 | 27916 | C. Thomas Tyler |
Fixed bug where verify_sdp.sh can wrongly give 'Current user ... does not own' error if started from a directory other than one owned by defined OSUSER. Fixed formatting error. Added quotes in 'case' per ShellCheck warning. |
||
#37 | 27750 | C. Thomas Tyler |
upgrade.sh v4.6.9: * Fixed issue where where patch-only upgrades of instances after the first in a multi-instance environment are skipped. * Corrected error message for scenario where downgrades are attempted; the logic was correct but error message was confusing. verify_sdp.sh v5.17.3: * Extended '-skip version' meaning to also skip new live binary version comparison checks. Related updates: * A call to verify_sdp.sh in the switch_db_files() function in backup_functions.sh now skips the version check. * A call to daily_checkpoint.sh now skips the version check. #review-27743 |
||
#36 | 27746 | C. Thomas Tyler |
verify_sdp.sh v5.17.2: * Downgraded more broker/proxy errors to warnings. * Added warning message for case where '-online' is specified and there is no live running p4p/p4broker (and they are defined as running run per existence of Instance Bin init scripts). |
||
#35 | 27745 | C. Thomas Tyler |
verify_sdp.sh v5.17.1: * Added missing check for 'p4 info' against proxy port. * New -online checks for p4broker/p4p downgraded to Warnings. |
||
#34 | 27742 | C. Thomas Tyler |
verify_sdp.sh v5.17.0: * Enhanced -online to also work for p4broker and p4p. * Added comparison check for live running vs. symlink'd version of servers (p4d, p4broker, and p4p). This is a part of the solution for SDP-625. * When checking for symlinks, also added checks that symlink target dirs exist. * For the special case where 'p4d' and 'p4p' are running on the same machine (according to the Instance Bin init scripts), proxy misconfiguration issues such as a missing /p4/N/cache symlink or a missing /hxdepots/p4/N/cache directory are reported as Warnings rather than Errors. |
||
#33 | 27737 | C. Thomas Tyler |
In verify_sdp.sh, made check for SDP password file a skippable check. This also skips the check for SDP_ADMIN_PASSWORD_FILE variable. See: SDP-641: Get SDP to work without a cleartext password file. #review-27732 |
||
#32 | 27731 | C. Thomas Tyler | Fixed logic bug that prevented check for existence of password file. | ||
#31 | 27723 | C. Thomas Tyler |
In verify_sdp.sh, handle scenario where there are no p4*_N.service files at all. Also, imrpove error messages to advise on necessary fix. |
||
#30 | 27722 | C. Thomas Tyler |
Refinements to @27712: * Resolved one out-of-date file (verify_sdp.sh). * Added missing adoc file for which HTML file had a change (WorkflowEnforcementTriggers.adoc). * Updated revdate/revnumber in *.adoc files. * Additional content updates in Server/Unix/p4/common/etc/cron.d/ReadMe.md. * Bumped version numbers on scripts with Version= def'n. * Generated HTML, PDF, and doc/gen files: - Most HTML and all PDF are generated using Makefiles that call an AsciiDoc utility. - HTML for Perl scripts is generated with pod2html. - doc/gen/*.man.txt files are generated with .../tools/gen_script_man_pages.sh. #review-27712 |
||
#29 | 27719 | C. Thomas Tyler |
Fixed issue preventing systemd service checks from working. Expanded scope of systemd checks to include any services related to the instance based on namving convention (p4*_N*.service), in addition to the checks for the 3 standard services (p4d_N/p4broker_N/p4p_N). This would include, for example, p4dtg_1. #review-27720 |
||
#28 | 27684 | C. Thomas Tyler |
Fixed invalid reporting of crontab issue if run on proxy host. #review-27685 |
||
#27 | 27412 | C. Thomas Tyler | Refined logic for test skipping. | ||
#26 | 27410 | C. Thomas Tyler | Added 'p4root' to list of tests that can be skipped. | ||
#25 | 27403 | C. Thomas Tyler |
verify_sdp.sh v5.14.0: * Added check for P4TRUST file if P4PORT starts with ssl:. * Added '-skip' option to avoid check for P4TICKETS and P4TRUST files. |
||
#24 | 27372 | C. Thomas Tyler | Added 'sort' to list of required utils in early preflight checks. | ||
#23 | 27367 | C. Thomas Tyler |
verify_sdp.sh v5.13.1: * Refined 'excess binaries in path' check ignore '.' * Dedupe 'excess binaries in path' via 'sort -u'. * Removed obsolete doc ref comment tags (non-functional change). #review-27368 |
||
#22 | 27348 | C. Thomas Tyler | Added option to skip checks requiring a healty offline_db. | ||
#21 | 27207 | C. Thomas Tyler |
upgrade.sh v4.3.0: * Greatly enhanced documentation and examples. * Before doing second journal rotation: - Added wait for 'p4 storage -w' (for 'to or thru' P4D 2019.1). - Added wait for 'p4 upgrades|grep -v completed' (for P4D 2020.2+) * Added check for whether p4broker and p4p were online at the start of upgrade processing. Only start those services that were running at the beginning of processing are now started after the binaries and symlinks are updated. For the broker, only the broker with the default configuration is stopped and started; DFM brokers are ignored by this script (thus making this script compaitble with using DFM brokers). * Fixed bug where '-c' (Protections table comment conversion) would have failed due to 'p4d' addition of 'Update:' field to the Protections table. Also generally enhanced logic to convert Protections table comments. * Added support for operation on proxy-only and broker-only hosts. Processing of upgrades for p4d occur only if /p4/N/bin/p4d_N_init script exists on the machine. * Refined lexigraphical P4D version comparsion checks. verify_sdp.sh v5.12.0: * Added support for proxy-only and broker-only hosts. The existence of a *_init script in the instance bin dir for any of the Helix server binaries p4d/p4p/p4broker indicate they are configured, determining what tests are executed or skipped. * Added check_file_x() function to check for execute bit on files. In backup_functions.sh, fixed is_server_up() to avoid displaying output. #review-27208 |
||
#20 | 27204 | C. Thomas Tyler | Fixed error where 'p4' client was reported as 'p4p' if found in PATH. | ||
#19 | 27191 | C. Thomas Tyler | Fixed typo. | ||
#18 | 27189 | C. Thomas Tyler |
verify_sdp.sh v5.11.0: * Added check for existance of $P4ROOT/server.id file. * Added optional check that P4MASTER_ID value starts with 'master'; this can be disabled with '-skip masterid'. #review-27190 |
||
#17 | 27170 | C. Thomas Tyler | Fixed typos in docs; no functional changes. | ||
#16 | 27169 | C. Thomas Tyler |
Fixed typo in error message. Thanks @roadkills_r_us! |
||
#15 | 27099 | C. Thomas Tyler |
Eliminated extra /p4/Version file. The standard for determining the SDP version for 2020.1 and forward will be the file /p4/sdp/Version. This is more clear. Job SDP-564 evolved slightly (per DevNotes in the job). The goal remains the same, to standardize the method of determining the SDP version for SDP 2020.1+. It was deemed that having an extra copy in /p4/Version will not help with that, and instead would introduce more failure modes and possibilities for out-of-sync files. This does mean the 'tarball extraction' sdp folder that is symlinked to from /p4/sdp is now a critical part of the SDP installation. This is normal and as documented, though there have been cases where SDP is copied from one machine to another in some incomplete way, e.g. rysnc of /p4/common but not /hxdepots/sdp and the symlink to it from /p4/sdp. However, the verify_sdp.sh will catch that form of misconfiguration. #review-27100 |
||
#14 | 27082 | C. Thomas Tyler |
verfiy_sdp.sh v5.9.2: Tweaked check for excess binaries in path to differentiate client (p4) from server (p4d/p4broker/p4p), Report excess servers as an error, and excess client as a warning (and thus still allowing a happy zero exit code). This is needed for compatibility with H4G. |
||
#13 | 27064 | C. Thomas Tyler |
Fixed issue where 'source p4_vars' hangs if load_checkpoint.sh is running. Added new semaphore file, $P4ROOT/P4ROOT_not_usable.txt. This is used in a way similar to 'offline_db_usable.txt' in the offline_db, except that this file only exists when the databases in P4ROOT are not usable. This is the opposite of how offline_db_usable.txt works, because P4ROOT is expected to be usable 99.9% fo the time. p4d_base will refuse to start p4d if this file exists, protecting against possible operator errors (like trying to start p4d when a checkpoint is still loading). Added check_file_dne() function to verify_sdp.sh to confirm a named file does not exist. Added checks in verify_sdp.sh that P4ROOT_not_usable.txt does not exist in P4ROOT or offline_db. Modified switch_db_files() (called by refresh_P4ROOT_from_offline_db.sh) to properly use the new P4ROOT_not_usable.txt safety file. Fixed bugs in p4d_base that could cause p4d_init.log to be overwritten if error output was generated. Removed call to 'backup_functions.sh' in p4d_base, as on balance it added more complexity than needed. #review-27065 |
||
#12 | 26982 | C. Thomas Tyler |
mkdirs.sh v4.1.0: * Accounted for directory structure change of Maintenance to Unsupported. * Added standard command line processing with '-h' and '-man' doc flags, and other flags (all documented). * Added in-code docs and updated AsciiDoc. * Enhanced '-test' mode to simulate /hx* mounts. * Enhanced preflight testing, and fixed '-test' mode installs. * Added support for installing to an alternate root directory. * Added '-s <ServerID>' option to override REPLICA_ID. * Added '-S <TargetServerID>' used for replicas of edge servers. * Added '-t <server_type>' option to override SERVER_TYPE. * Added '-M' option to override mount points. * Added '-f' fast option to skip big chown/chmod commands, and moved those commands near the end as well. verify_sdp.sh v5.9.0: * Added check for /p4/Version file, and checked that other legacy SDP methods of checking version * Added sanity check for crontab. * Added 'test skip' mechanism to skip certain tests: - crontab: Skip crontab check. Use this if you do not expect crontab to be configured, perhaps if a different scheduler is used. - license: Skip license related checks. - version: Skip version checks. - excess: Skip checks for excess copies of p4d/p4p/p4broker in PATH. * Added VERIFY_SDP_SKIP_TEST_LIST setting ton instance_vars.template, to define a standard way to have verify_sdp.sh always skip certain tests for a site. * Extended '-online' checks to check for bogus P4MASTERPORT, a common config error. Update test_SDP.py: * Adjusted test suite to account for various changes in mkdirs.sh. * Added 'dir' parameter to run_cmd() and sudo_cmd(), to run a command from a specified directory (as required to test new mkdirs.sh) * Added check_links() similar to existing check_dirs() function. === Upgrade Process Changes === Made /p4/common/bin/p4d/p4/p4broker/p4p shell script rather than binary. This changes the way SDP new binaries are staged for upgrade. For safety, exes are now staged to a director outside the PATH, the /p4/sdp/exes folder. A new 'get_latest_exes.sh' script simplifies the task of pulling executables from the Perforce FTP server. This can be used 'as is' for environments with outbound internet access, and is useful in any case to describe now to acquire binaries. This addresses an issue where a p4d binary staged for a future upgrade might be called before the actual upgrade is performed. upgrade.sh v4.0.0: * All preflight checks are now done first. Added '-p' to abort after preflight. * Added '-n' to show what would be done before anything is executed. * Minimalist logic to start/stop only servers that are upgrade, and apply upgrades only as needed. * Staging of exes for upgrade is now separate from /p4/common/bin * Improved in-code docs, added '-h' and '-man' options. * Retained pre/post P4D 2019.1 upgrade logic. |
||
#11 | 26814 | C. Thomas Tyler |
Skip license check on platforms for which 'date' utility is incompatible (e.g. Mac). |
||
#10 | 26718 | Robert Cowham |
Rename P4MASTER to P4MASTERHOST for clarity with comments in: - mkdirs.cfg/mkdirs.sh - p4_<instance>.vars - other files which reference Remove unnecessary sed for p4p.template |
||
#9 | 26637 | Robert Cowham |
Include script help within doc Requires a couple of tags in the scripts themselves. |
||
#8 | 26422 | C. Thomas Tyler |
Fixed issue where checks cause db.config to be created in otherwise empty P4ROOT. Fixed typo in declaration. |
||
#7 | 26400 | C. Thomas Tyler |
Added refresh_P4ROOT_from_offline_db.sh. Updated backup_functions.sh to support functionality for db refresh. Upgrade start_p4d() and stop_p4d() to use systemd if available, else use the underlying SysV init scripts. Updated verify_sdp.sh to be called from other scripts (sans its own logging). Added many checks to verify_sdp.sh to support P4ROOT/offline_db swap. Logic in P4ROOT/offline_db swap is more careful about what gets swapped. Added start_p4broker() and stop_p4broker() that behave similarly. More shellcheck compliance. #review-26401 |
||
#6 | 26387 | C. Thomas Tyler |
verify_sdp.sh v5.5.1: * Removed a redundant check. * Cleaned up some output issues. |
||
#5 | 26385 | C. Thomas Tyler |
Added license expiration check. Added new 'warning' indicator, so some things may be reported as warnings rather than errors. Fixed bug in not-released dev version handling su/exec when started as root. Enhanced verify_sdp.sh to support being called by other scripts by giving a reliable exit code. Errors detected result in a non-zero exit code; warnings do not. This is in support of SDP-444, but does not complete that job. Maintained shellcheck compliance. #review-26386 |
||
#4 | 26074 | C. Thomas Tyler |
Added missing flags in usage (-h) showing in '-man'. No functional change. |
||
#3 | 25372 | C. Thomas Tyler |
Various updates in verify_sdp.sh v5.3.0: * Passes shellcheck v.0.6.0 check; which drove various minor code changes, e.g. using bash built-in 'command -v' rather than system utility 'which', and others. * Improved usage of mktemp util to use P4TMP in a template, to avoid SELinux issues. * Fixed issue where early calls to bail() resulted in output not being displayed. * Bullet-proofed for scenarios where script is called by the wrong user, or called by correct user after first calling by wrong user and log ownership is wrong, or user can't write log file, or P4TMP not defined, etc. * Clarified TODO comments. |
||
#2 | 25206 | C. Thomas Tyler |
Removed logic that uses 'p4d -cset' to force the value for P4JOURNAL, and also automatic journal rotation on server startup. Added related logic to verify_sdp.sh to ensure there is one source of truth for the P4JOURNAL definition. === On Journal Rotation at Server Startup === The goal with journal rotation on server stratup is noble, to make it so any potential journal corruption *always* appears at the end of a numbered journal file, rather than being in the middle of the active journal. This can make it easier and faster to recover from journal corruption caused by sudden power loss, kernel panic, a p4d bug/crash, etc. However, the implementation causes problems (noted below). === On Forcing P4JOURNAL === The goal of forcing the value of P4JOURNAL via db.config is also noble, in that having a value anything other than the SDP standard can really wreak havoc with things. This is generally not an issue in a 'fresh' SDP install, but can be an issue (wreak havoc!) in cases where 'p4 configure' was used to set a value for P4JOURNAL that conflicts with the value defined by the SDP environment mechanism, which is in turn passed to 'p4d' on the command line. Even if the value defined differently, it should be set in to exactly one value, and exactly one mechanism. The current implementation causes problems (noted below). == Problems with setting P4JOURNAL in db.config == 1. Things Break The forced P4JOURNAL set via 'p4d -cset' causes a mild form of journal corruption that breaks 'standby' replicas using journalcopy, as this type of replica is extremely sensitive to the contents of every byte in the journal file, and doesn't allow for use of 'p4d -cset' to modify the P4JOURNAL file. While it does not cause any actual loss of data, it does require manual reset to fix things. In the case of a site-wide topology with a mandatory standby replica, it causes global replication to stall. 2. Not our Place (not the place of SDP scripts) Based on the above and taking a step back, I think this script behavior of forcing a back-door journal rotation is simply too intrusive for what SDP scritps should be allowed to do. They live to have some understanding of p4d workings, but shoulnd't pretend to have too much insight into the inner workings of p4d. == Problem with Always-On Journal Rotation on Start == 1. What the wah? This confuses admins by incrementing the journal counter unexpectedly. In Battle School training classes, for example, students (Perforce admins) are confused by seemingly random journal incrementing. While this could be documented and trained for, it violates the principal of least surprise, and is not typical 'p4d' beavhior. 2. Always vs. Rare It rotates the journal even when there is no corruption, which of course 99.99999% or more of the time at any given site. Anyone who has been through a corruption scenario is happy to have the corruption at the end rather than in the middle of a journal file -- as noted, the intent here is noble. But before we do any journal rotations, we should detect whether there is corruption. Turns out we have a means to detect journal corruption at the end of the current/active journal file, and should employ such detection and handle it in some approrpaite manner, e.g. by expanding the 'force_start' logic in this p4d_base init script. Journal corrption detection and preliminary handling may be added in a future SDP release. When the journal is truly corrupted, global replication will stall in any case, so measure like journal file rotation may be called for in that scenario. 3. Accelerated Deletion of Backups Increased journal counter rotations result in unexpectedly fast removal of backups. Admins are used to thinking that roughly, "one journal rotation is roughly one day." Settings like KEEPLOGS, KEEPCKPS, and KEEPJNLS trigger off the number of journal rotatations, not the number of actual calendar days. Now, I think it's OK that journal rotations and days don't match precisely. In a typical "big deal" maintenance window, for example, there might be an additional 1-3 journal rotations induced by extra checkpoints or journals being created over the course of a maintenance activity. But there may be a dozen more 'p4d' restarts during sanity testing and playing around with things. With the current logic, each restart causes another journal rotation. By the end fo the weekend, your next call to daily_checkpoint might remove more of your recent backups than you'd like or expect. (A long standing safety feature always preserves the last few, but still we don't want to delete more than desired.) === Foor for Thougt: KEEP* = numbrer of days? === Making it so KEEPLOGS/KEEPJNLS/KEEPCKPS mean literally number of days rather than journal rotations is worthy of consideration. That's beyond the scope of this change though. #review @robert_cowham @josh |
||
#1 | 24804 | C. Thomas Tyler |
Terminology tweak, 'validate' -> 'verify'. #review @robert_cowham |
||
//guest/perforce_software/sdp/dev/Server/Unix/p4/common/bin/validate_sdp.sh | |||||
#3 | 24534 | C. Thomas Tyler |
Various enhancements and internal refactoring for validate_sdp.sh. Added important check that /p4/N is a dir, not a symlink. #review-24532 @robert_cowham |
||
#2 | 24356 | C. Thomas Tyler |
Enhancements to validate_sdp.sh: * Added simple bold ALL CLEAN message to look for. * Added check_env_var() function to check shell environment variables, with some calls to it. * Added check_configurable() to check for configurables. This is implemented using back-door check methodology (using 'p4d_N -cshow') so values can be checked with P4D offline. This replaced stub function check_var(). * Removed stub function check_configurables(). It's easier to understand if all checks in the Main section of the code. * Changed so checks requiring p4d to be online are not done by default; added '-online' flag to run those tests. This is because I anticpate typical usage of the validator to be requiring it to be report ALL CLEAN before starting P4D after a server upgrade. * Added check for new $SDP_ADMIN_PASSWORD_FILE variable. * Added check admin password file pointed to by $SDP_ADMIN_PASSWORD_FILE. * Added errmsg() function, with corresponding tweak to bail(). * Consolidated Log an LOGIFLE to just LOGFILE. * Removed a few items from TOOD comments that got done. * Made a few tweaks for style normalization: - Functions are lowercase with undescore separators. - Functions vars are lowercase-initiated camelCase. - Indentation: 3 spaces for functions/loops/etc. * Added run() function replacing cmd() stub function. * Enhanced p4login check. * Added comment noting why this script uses self-contained copies of functions defined in other SDP files in /p4/common/lib. * And other things. Warning: In the short run, this may fail tests as the new SDP_ADMIN_PASSWORD_FILE variable is also pending review. #review @robert_cowham |
||
#1 | 23640 | Robert Cowham | Super basic validation - placeholder for many more tests to come! |