#!/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 SDP_INSTANCE="${SDP_INSTANCE:-UnsetSDPInstance}" declare Version=5.18.2 declare ThisScript="${0##*/}" declare SDPInstallRoot="/p4" declare SDPCommon="$SDPInstallRoot/common" declare SDPCommonBin="$SDPCommon/bin" declare SDPCommonCfg="$SDPCommon/config" declare SDPEnvFile="$SDPCommonBin/p4_vars" declare SDPOwner= declare CmdArgs="$*" declare CmdLine="$0 $CmdArgs" declare SDPVersionA= declare SDPVersionB= declare SDPVersionC= declare -i ServerOnline=0 declare -i ErrorCount=0 declare -i WarningCount=0 declare -i CheckCount=0 declare -i SilentMode=0 declare -i DoCrontabTest=1 declare -i DoCrontabTestWarn=0 declare -i DoLicenseTest=1 declare -i DoLicenseTestWarn=0 declare -i DoVersionTest=1 declare -i DoVersionTestWarn=0 declare -i DoExcessBinaryTest=1 declare -i DoExcessBinaryTestWarn=0 declare -i DoMasterIDTest=1 declare -i DoMasterIDTestWarn=0 declare -i DoOfflineDBTest=1 declare -i DoOfflineDBTestWarn=0 declare -i DoP4ROOTTest=1 declare -i DoP4ROOTTestWarn=0 declare -i DoPasswordChecks=1 declare -i DoPasswordChecksWarn=0 declare -i DoP4TFilesTest=1 declare -i DoP4TFilesTestWarn=0 declare -i ExcessServerBinariesFound=0 declare -i P4DServer=0 declare -i P4BrokerServer=0 declare -i P4ProxyServer=0 declare SkipTestList= declare WarnTestList= declare BadLog= declare ThisUser= declare LicenseInfo= declare LicenseExpiration= declare CurrentTime= declare ExpirationTime= declare TimeDiff= declare DaysDiff= declare LicenseDaysExpirationAlert=21 declare LinkP4ROOT= declare LinkOfflineDB= declare H1="==============================================================================" declare H2="------------------------------------------------------------------------------" declare Log="Unset" export P4TMP="Unset" declare TmpFile= #============================================================================== # Local Functions # Note: This script does not use SDP library files, as its purpose is to # verify the integrity of an SDP installation. Thus, it has its own # self-contained versions of some functions for which similar versions # would normally be sourced from files in /p4/common/lib, like libcore.sh. function msg () { echo -e "$*"; } 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 # error, usually $1 should be -h so that the longer usage message doesn't # obscure the error message. # # Sample Usage: # usage # usage -man # usage -h "Incorrect command line usage." # # This last example generates a usage error message followed by the short # '-h' usage summary. #------------------------------------------------------------------------------ function usage { declare style=${1:--h} declare errorMessage=${2:-Unset} if [[ $errorMessage != Unset ]]; then echo -e "\\n\\nUsage Error:\\n\\n$errorMessage\\n\\n" >&2 fi echo "USAGE for verify_sdp.sh v$Version: verify_sdp.sh [<instance>] [-online] [-skip <test>[,<test2>,...]] [-warn <test>[,<test2>,...]] [-si] [-L <log>|off ] [-D] or verify_sdp.sh -h|-man " if [[ $style == -man ]]; then 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 requires p4d, p4broker, and/or p4p to be online. Any servers for which there are *_init scripts in the Instance Bin directory must be online. The Instance Bin directory is the /p4/N/bin directory, where N is the SDP instance name. -skip <test>[,<test2>,...] Specify a comma-delimited list of named tests to skip. Valid test names are: * cron|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. * masterid: Skip check ensuring ServerID of master starts with 'master'. * offline_db: Skip checks that require a healthy offline_db. * p4root: Skip checks that require healthy P4ROOT db files. * p4t_files: Skip checks for existence of P4TICKETS and P4TRUST files. * passwd|password: Skip SDP password checks. * version: Skip version checks. As an alternative to using the '-skip' option, the shell environment variable VERIFY_SDP_SKIP_TEST_LIST can be set to a comma-separated list of named tests 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 $SDPCommonCfg/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 -warn <test>[,<test2>,...] Specify a comma-delimited list of named tests that will be reported as warnings rather than errors. The list of valid test names as the same as for the '-skip' option. As an alternative to using the '-warn' option, the shell environment variable VERIFY_SDP_WARN_TEST_LIST can be set to a comma-separated list of name tests to skip. Using the command line parameter is the best choice for temporarily converting errors to warnings, while specifying the environment variable is better for making the conversion to warnings permanent. The variable should be set in $SDPCommonCfg/p4_N.vars file. If the '-warn' option is provided, the VERIFY_SDP_WARN_TEST_LIST variable is ignored (not appended to). So it may make sense to reference the variable on the command line. For example, if the value of the variable is 'crontab', to convert to warnings for crontab and excess binaries tests, you could specify: -warn \$VERIFY_SDP_WARN_TEST_LIST,excess -si Silent mode, useful for cron operation. Both stdout and stderr are still captured in the log. The '-si' option cannot be used with '-L off'. -L <log> Specify the log file to use. The default is /p4/N/logs/verify_sdp.log The special value 'off' disables logging to a file. Note that '-L off' and '-si' are mutually exclusive. -D Set extreme debugging verbosity. HELP OPTIONS: -h Display short help message -man Display man-style help message EXAMPLES: Example 1: Typical usage: This script is typically called after SDP update with only the instance name or number as an argument, e.g.: verify_sdp.sh 1 Example 2: Skipping some checks. verify_sdp.sh 1 -skip crontab Example 3: Automation Usage If used from automation already doing its own logging, use -L off: verify_sdp.sh 1 -L off LOGGING: This script generates a log file and also displays it to stdout at the end of processing. By default, the log is: /p4/N/logs/verify_sdp.log. The exception is usage errors, which result an error being sent to stderr followed usage info on stdout, followed by an immediate exit. If the '-si' (silent) flag is used, the log is generated, but its contents are not displayed to stdout at the end of processing. EXIT CODES: An exit code of 0 indicates no errors were encountered attempting to perform verifications, and that all checks verified cleanly. " fi exit 1 } #------------------------------------------------------------------------------ # Function: terminate function terminate { # Disable signal trapping. trap - EXIT SIGINT SIGTERM [[ "$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 sort 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 $SDPCommonBin" CheckCount+=1 if cd "$SDPCommonBin"; then cd "$OLDPWD" || bail "Failed to cd to $OLDPWD. Aborting." else errmsg "Could not cd to: $SDPCommonBin" return 1 fi msg "Verified: cd works to: $SDPCommonBin" msg "Preflight Check: Checking current user owns $SDPCommonBin" # shellcheck disable=SC2012 SDPOwner="$(ls -ld "$SDPCommonBin" | awk '{print $3}')" ThisUser="$(id -n -u)" CheckCount+=1 if [[ "$ThisUser" == "$SDPOwner" ]]; then msg "Verified: Current user [$ThisUser] owns $SDPCommonBin" else errmsg "Current user [$ThisUser] does not own $SDPCommonBin. This most likely means this script is running as the wrong user. It could also mean the $SDPCommonBin directory is not owned by the correct owner, which should be the OS account under which the p4d process runs." return 1 fi msg "Preflight Check: Checking /p4 and /p4/<instance> are local dirs." if ! check_local_instance_home_dir "$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. # # Inputs: # $1 - $file path to check. Required. # $2 - Optional error message to display if file is missing. # Default: "Missing file [$file]." # $3 - $warningOnly. If 0, an error is displayed if the file does not exist. # If 1, a warning is displayed instead of an error. Default is 0. # # 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 -i 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. # # Inputs: # $1 - $file path to check. Required. # $2 - Optional error message to display if file is missing. # Default: "Missing not executable [$file]." # $3 - $warningOnly. If 0, an error is displayed if the file does not exist. # If 1, a warning is displayed instead of an error. Default is 0. # # 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 -i warningOnly=${3:-0} CheckCount+=1 msg "Checking executable file $file" [[ -x "$file" ]] && return 0 if [[ "$warningOnly" -eq 0 ]]; then errmsg "$errMsg: [$file]." else warnmsg "$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(). # # Inputs: # $1 - $file path to check. Required. # $2 - Optional error message to display if file exists. # Default: "This file should not exist: [$file]." # $3 - $warningOnly. If 0, an error is displayed if the file does not exist. # If 1, a warning is displayed instead of an error. Default is 0. # # # 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 -i warningOnly=${3:-0} CheckCount+=1 msg "Confirming this file does not exist: $file" [[ ! -f "$file" ]] && return 0 if [[ "$warningOnly" -eq 0 ]]; then errmsg "$errMsg: [$file]." else warnmsg "$errMsg: [$file]." fi return 1 } #------------------------------------------------------------------------------ # Function: check_configurable ($instance, $configurable, $scope, $expectedVal, $errMsg1, $errMsg2, $warningOnly) # # 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. # $7 - $warningOnly. If 0, an error is displayed if the configurable is not # defined or does not have the expected value. # If 1, a warning is displayed instead of an error. Default is 0. # # Return Codes: # 1 - Verifications failed. # 0 - Verifications passed. # # Sample Usage: # check_configurable "$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="${6:-Value does not match what is expected}" local -i warningOnly=${7:-0} 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 ${configurable} is defined at least once." fi else if [[ "$warningOnly" -eq 0 ]]; then errmsg "$errMsgMissing for configurable [${scope}:${configurable}]." else warnmsg "$errMsgMissing for configurable [${scope}:${configurable}]." fi return 1 fi else if [[ -n "$value" ]]; then if [[ "$scope" != "ALL" ]]; then if [[ "$warningOnly" -eq 0 ]]; then errmsg "Configurable ${configurable} should not be set with 'p4 configure set' but has a value for ServerID ${scope} of: ${value}" else warnmsg "Configurable ${configurable} should not be set with 'p4 configure set' but has a value for ServerID ${scope} of: ${value}" fi return 1 else if [[ "$warningOnly" -eq 0 ]]; then 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)." else warnmsg "Configurable ${configurable} should not be set with 'p4 configure set' but has a value for ServerID ${detectedScope} of: ${value} (and possibly for other ServerIDs)." fi return 1 fi else if [[ "$scope" != "ALL" ]]; then msg "Verified: Configurable ${scope}:${configurable} is undefined (as expected)." else msg "Verified: Configurable ${configurable} is undefined (as expected) for all ServerID values." fi fi fi [[ "$expectedValue" == "NoExpectedValue" ]] && return 0 CheckCount+=1 if [[ "$expectedValue" != "UNDEF" ]]; then if [[ "$value" == "$expectedValue" ]]; then msg "Verified: Configurable ${scope}:${configurable} has expected value [$value]." else if [[ "$warningOnly" -eq 0 ]]; then errmsg "$errMsgBadValue for variable [${scope}:${configurable}]\\n\\tExpected value: [$expectedValue]\\n\\tActual value: [$value]" else warnmsg "$errMsgBadValue for variable [${scope}:${configurable}]\\n\\tExpected value: [$expectedValue]\\n\\tActual value: [$value]" fi return 1 fi fi return 0 } #------------------------------------------------------------------------------ # Function: check_env_var ($instance, $var, $expectedval, $msg1, $msg2, $warningOnly) # # 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. # $6 - $warningOnly. If 0, an error is displayed if the environment variable # is not set or does not match the expected value. # If 1, a warning is displayed instead of an error. Default is 0. # # 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 -i warningOnly=${6:-0} local value= CheckCount+=1 eval unset "${var}" # shellcheck disable=SC1090 source "$SDPEnvFile" "$instance" set +u if [[ -n "$(eval echo \$"${var}")" ]]; then msg "Verified: Variable ${var} is defined." set -u else if [[ "$warningOnly" -eq 0 ]]; then errmsg "$errMsgMissing for variable [$var]." else warnmsg "$errMsgMissing for variable [$var]." fi 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 if [[ "$warningOnly" -eq 0 ]]; then errmsg "$errMsgBadValue for variable [$var]\\n\\tExpected value: [$expectedValue]\\n\\tActual value: [$value]" else warnmsg "$errMsgBadValue for variable [$var]\\n\\tExpected value: [$expectedValue]\\n\\tActual value: [$value]" fi return 1 fi return 0 } #------------------------------------------------------------------------------ # Function: check_local_instance_home_dir ($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/or 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;; (-warn) WarnTestList="$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 $SDPEnvFile <instance>\\n\\nor by passing in the instance name as a parameter to this script.\\n" [[ "$SilentMode" -eq 1 && "$Log" == "off" ]] && \ usage -h "Cannot use '-si' with '-L off'." #============================================================================== # Main Program # shellcheck disable=SC1090 source "$SDPEnvFile" "$SDP_INSTANCE" ||\ bail "Failed to load SDP environment for instance $SDP_INSTANCE." # shellcheck disable=SC1090 disable=SC1091 source "$P4CBIN/backup_functions.sh" ||\ bail "Failed to load backup_functions.sh." [[ "${OSUSER:-Unset}" == "Unset" ]] &&\ bail "The critical OSUSER setting is not defined in $SDPEnvFile. 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 "$SDPCommonBin/${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 [[ -z "$WarnTestList" && -n "${VERIFY_SDP_WARN_TEST_LIST:-}" ]]; then WarnTestList="$VERIFY_SDP_WARN_TEST_LIST" fi if [[ -n "$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;; (p4root) DoP4ROOTTest=0;; (p4t_files) DoP4TFilesTest=0;; (passwd|password) DoPasswordChecks=0;; (version) DoVersionTest=0;; (*) errmsg "Invalid test name specified with '-skip': $test";; esac done fi if [[ -n "$WarnTestList" ]]; then for test in $(echo "$WarnTestList" | tr ',' ' '); do case "$test" in (cron|crontab) DoCrontabTestWarn=1;; (excess) DoExcessBinaryTestWarn=1;; (license) DoLicenseTestWarn=1;; (masterid) DoMasterIDTestWarn=1;; (offline_db) DoOfflineDBTestWarn=1;; (p4root) DoP4ROOTTestWarn=1;; (p4t_files) DoP4TFilesTestWarn=1;; (passwd|password) DoPasswordChecksWarn=1;; (version) DoVersionTestWarn=1;; (*) errmsg "Invalid test name specified with '-warn': $test";; esac done fi # Logs should be defined to /p4/N/logs after sourcing the environment # file above; default to /tmp for cases of incomplete environment where # LOGS is not defined. export LOGS="${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 "$SDP_INSTANCE" SERVERID if [[ "$DoPasswordChecks" -eq 1 ]]; then if check_env_var "$SDP_INSTANCE" SDP_ADMIN_PASSWORD_FILE "$P4CCFG/.p4passwd.$P4SERVER.admin" $DoPasswordChecksWarn; then check_file "$SDP_ADMIN_PASSWORD_FILE" "SDP admin password file doesn't exist" $DoPasswordChecksWarn fi else msg "Skipping password file check per '-skip'." fi msg "${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 [[ $((P4DServer+P4BrokerServer+P4ProxyServer)) -eq 0 ]]; then CheckCount+=1 errmsg "No servers (p4d, p4p, p4broker) are configured." fi if [[ -d "/etc/systemd/system" ]]; then msg "${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##*/}'.\\nThis can be fixed with:\\n\\tsudo chmod 644 $P4DSystemdServiceFile\\n" fi fi # shellcheck disable=SC2154 if [[ -e "$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##*/}'.\\nThis can be fixed with:\\n\\tsudo chmod 644 $P4BrokerSystemdServiceFile\\n" fi fi # shellcheck disable=SC2154 if [[ -e "$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##*/}'.\\nThis can be fixed with:\\n\\tsudo chmod 644 $P4ProxySystemdServiceFile\\n" fi fi # Check for any other services related to this instance by naming # convention, skipping the standard 3 (p4d/p4broker/p4p) already checked # above. for svcFile in /etc/systemd/system/p4*_"${SDP_INSTANCE}"*.service; do case "$svcFile" in ("$P4DSystemdServiceFile") continue;; ("$P4BrokerSystemdServiceFile") continue;; ("$P4ProxySystemdServiceFile") continue;; # If $svcFile contains a literal '*', there are no p4*_N*.service files. ('/etc/systemd/system/p4*_'*) continue;; (*) svcName="${svcFile##*/}" svcName="${svcName%.service}" CheckCount+=1 if systemctl cat "$svcName" > /dev/null 2>&1; then msg "Verified: Permissions OK to status w/o sudo for service $svcName." else errmsg "Could not run 'systemctl status $svcName'.\\nThis can be fixed with:\\n\\tsudo chmod 644 $svcFile\\n" fi ;; esac done fi if [[ "$P4DServer" -eq 1 ]]; then msg "${H2}\\nChecking for a few database files." # Check whether P4ROOT db files exist if [[ "$DoP4ROOTTest" -eq 1 ]]; then for file in db.counters db.domain db.config; do check_file "$P4ROOT/$file" "Expected database file doesn't exist." $DoP4ROOTTestWarn done fi # Check whether offline_db files exist if [[ "$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." $DoOfflineDBTestWarn done fi fi msg "${H2}\\nChecking for existence of key files." check_file "$P4BIN" "The p4 binary (or symlink) doesn't exist." if [[ "$DoP4TFilesTest" -eq 1 ]]; then check_file "$P4TICKETS" "The P4TICKETS file doesn't exist." \ $DoP4TFilesTestWarn if [[ "$P4PORT" == "ssl:"* ]]; then check_file "$P4TRUST" "The P4TRUST file doesn't exist and SSL is enabled." \ $DoP4TFilesTestWarn fi fi if [[ "$P4DServer" -eq 1 ]]; then if [[ "$DoP4ROOTTest" -eq 1 ]]; then check_file "$P4ROOT/server.id" \ "The required $P4ROOT/server.id file is missing." $DoP4ROOTTestWarn check_file_dne "$P4ROOT/P4ROOT_not_usable.txt" \ "P4ROOT is not in a usable state." $DoP4ROOTTestWarn fi if [[ "$DoOfflineDBTest" -eq 1 ]]; then check_file "$OFFLINE_DB/offline_db_usable.txt" \ "Offline database not in a usable state." $DoOfflineDBTestWarn check_file_dne "$OFFLINE_DB/P4ROOT_not_usable.txt" \ "Offline database has P4ROOT_not_usable.txt file." $DoOfflineDBTestWarn fi if [[ "$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 if [[ "$DoLicenseTestWarn" -eq 1 ]]; then warnmsg "The license is expired." else errmsg "The license is expired." fi elif [[ "$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 "$LicenseDaysExpirationAlert" ]]; then if [[ "$DoLicenseTestWarn" -eq 1 ]]; then warnmsg "License will expire within $LicenseDaysExpirationAlert days." else errmsg "License will expire within $LicenseDaysExpirationAlert days." fi fi else msg "Skipping license check due to incompatible 'date' utility on this OS." fi elif [[ "$LicenseInfo" == *"support ends"* ]]; then msg "Info: License is perpetual." fi else if [[ "$DoLicenseTestWarn" -eq 1 ]]; then warnmsg "Could not determine license info from license file." else errmsg "Could not determine license info from license file." fi fi fi else msg "Skipping license check per '-skip'." fi fi if [[ "$P4DServer" -eq 1 ]]; then msg "${H2}\\nChecking configurables values." msg "This next test checks that P4JOURNAL is defined only as a shell environment variable and is not set in db.config.\\n" check_configurable "$SDP_INSTANCE" P4JOURNAL ALL UNDEF check_configurable "$SDP_INSTANCE" journalPrefix any "$CHECKPOINTS/$P4SERVER" check_configurable "$SDP_INSTANCE" server.depot.root any "$DEPOTS" fi if [[ "$P4DServer" -eq 1 ]]; then if [[ "$DoP4ROOTTest" -eq 1 ]]; then msg "${H2}\\nChecking P4ROOT symlinks." 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 fi if [[ "$DoOfflineDBTest" -eq 1 ]]; then msg "${H2}\\nChecking offline_db symlinks." 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 if [[ "$DoP4ROOTTest" -eq 1 && "$DoOfflineDBTest" -eq 1 ]]; then 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 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." CheckCount+=1 targetDir=$(readlink "$link") if [[ -d "$targetDir" ]]; then msg "Verified: Target for $PWD/$link is a directory." else # If we are on a server that runs both 'p4p' and 'p4d', report proxy config # issues as Warnings rather than Errors. if [[ "$P4DServer" -eq 1 && "$link" == "cache" ]]; then warnmsg "Target for symlink $PWD/$link, $targetDir, is not a directory." else errmsg "Target for symlink $PWD/$link, $targetDir, is not a directory." fi fi else # If we are on a server that runs both 'p4p' and 'p4d', report proxy config # issues as Warnings rather than Errors. if [[ "$P4DServer" -eq 1 && "$link" == "cache" ]]; then warnmsg "$PWD/$link is not a symlink." else errmsg "$PWD/$link is not a symlink." fi fi done CheckCount+=1 if [[ -L "$SDPCommon" ]]; then msg "Verified: $SDPCommon is a symlink." else errmsg "Path $SDPCommon 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 "${H2}\\nDoing online checks." if [[ "$P4DServer" -eq 1 ]]; then 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 $P4MASTERPORT could not be verified." fi if [[ "$DoVersionTest" -eq 1 ]]; then CheckCount+=1 LinkP4DVersion=$("$P4DBIN" -V | grep ^Rev) LinkP4DVersion=${LinkP4DVersion#Rev. } LinkP4DVersion=${LinkP4DVersion%\.} LiveP4DVersion=$("$P4BIN" -ztag -F %serverVersion% info -s) if [[ "$LinkP4DVersion" == "$LiveP4DVersion" ]]; then msg "Verified: Live running p4d version matches expected version [$LinkP4DVersion]." else if [[ "$DoVersionTestWarn" -eq 0 ]]; then errmsg "Live running p4d version [$LiveP4DVersion] does not match expected version [$LinkP4DVersion]." else warnmsg "Live running p4d version [$LiveP4DVersion] does not match expected version [$LinkP4DVersion]." fi fi else msg "Skipping live p4d version check per '-skip'." fi fi if [[ "$P4BrokerServer" -eq 1 ]]; then CheckCount+=1 if run "$P4BIN -s -p $P4BROKERPORT info -s" "Checking 'p4 -s -p $P4BROKERPORT info -s'" 1; then msg "Verified: 'p4 -s -p $P4BROKERPORT info -s' output is nominal." else warnmsg "Could not verify the p4broker server is online (P4PORT=$P4BROKERPORT)." fi if [[ "$DoVersionTest" -eq 1 ]]; then CheckCount+=1 LinkP4BrokerVersion=$("$P4BROKERBIN" -V | grep ^Rev) LinkP4BrokerVersion=${LinkP4BrokerVersion#Rev. } LinkP4BrokerVersion=${LinkP4BrokerVersion%% (*} LiveP4BrokerVersion=$("$P4BIN" -p "$P4BROKERPORT" -ztag -F %brokerVersion% info -s) if [[ "$LinkP4BrokerVersion" == "$LiveP4BrokerVersion" ]]; then msg "Verified: Live running p4broker version matches expected version [$LinkP4BrokerVersion]." else if [[ -n "$LiveP4BrokerVersion" ]]; then warnmsg "Live running p4broker version [$LiveP4BrokerVersion] does not match expected version [$LinkP4BrokerVersion]." else warnmsg "Could not determine live running p4broker version and '-online' was specified." fi fi else msg "Skipping live p4broker version check per '-skip'." fi fi if [[ "$P4ProxyServer" -eq 1 ]]; then if run "$P4BIN -s -p $PROXY_PORT info -s" "Checking 'p4 -s -p $PROXY_PORT info -s'" 1; then msg "Verified: 'p4 -s -p $PROXY_PORT info -s' output is nominal." else warnmsg "Could not verify the p4p server is online (P4PORT=$PROXY_PORT)." fi if [[ "$DoVersionTest" -eq 1 ]]; then CheckCount+=1 LinkP4ProxyVersion=$("$P4PBIN" -V | grep ^Rev) LinkP4ProxyVersion=${LinkP4ProxyVersion#Rev. } LinkP4ProxyVersion=${LinkP4ProxyVersion%\.} LiveP4ProxyVersion=$("$P4BIN" -p "$PROXY_PORT" -ztag -F %proxyVersion% info -s) if [[ "$LinkP4ProxyVersion" == "$LiveP4ProxyVersion" ]]; then msg "Verified: Live running p4p version matches expected version [$LinkP4ProxyVersion]." else if [[ -n "$LiveP4ProxyVersion" ]]; then warnmsg "Live running p4p version [$LiveP4ProxyVersion] does not match expected version [$LinkP4ProxyVersion]." else warnmsg "Could not determine live running p4p version and '-online' was specified." fi fi else msg "Skipping live p4p version check per '-skip'." fi 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|p4pcm.pl) ' "$TmpFile"; then msg "Verified: Crontab for user $USER passed basic sanity check." elif [[ "$DoCrontabTestWarn" -eq 1 ]]; then warnmsg "Crontab for user $USER did not pass basic sanity check; missing call to daily_checkpoint.sh or replica_status.sh" 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 $SDPEnvFile matches." else errmsg "SDP Version from $SDPEnvFile isn't set or doesn't match /p4/sdp/Version:\\n[$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_VERSION' is not set. To fix:\\n\\tsource /p4/common/bin/p4_vars $SDP_INSTANCE\\n\\tp4 counter SDP_VERSION \"\$(cat /p4/sdp/Version)\"\\n\\tp4 counter SDP_DATE \"\$(date +'%Y-%m-%d')\"" else errmsg "SDP Version from 'p4 counter SDP_VERSION' doesn't match /p4/sdp/Version:\\n[$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 $SDPCommonBin 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 # shellcheck disable=SC2230 for exeInPath in $(which -a $exe | sort -u); do dir="${exeInPath%/*}" if [[ "$dir" != *"$P4CBIN" && "$dir" != "." ]]; then if [[ "$DoExcessBinaryTestWarn" -eq 1 ]]; then warnmsg "Executable $exe found outside $SDPCommonBin in PATH: $dir" ExcessServerBinariesFound=1 else errmsg "Executable $exe found outside $SDPCommonBin in PATH: $dir" ExcessServerBinariesFound=1 fi fi done done # Check for excess Helix client binary in PATH, and classify that as a warning. # shellcheck disable=SC2230 for exeInPath in $(which -a p4 | sort -u); do if [[ "$dir" != *"$P4CBIN" && "$dir" != "." ]]; then warnmsg "Executable 'p4' found outside $SDPCommonBin in PATH: ${exeInPath%/*}" fi done if [[ "$ExcessServerBinariesFound" -eq 0 ]]; then msg "Verified: No excess Helix server binaries found outside $SDPCommonBin." fi else msg "Skipping excess binary checks per '-skip'." fi if [[ "$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 if [[ "$DoMasterIDTestWarn" -eq 1 ]]; then warnmsg "The P4MASTER_ID value ($P4MASTER_ID) does not start with 'master'." else errmsg "The P4MASTER_ID value ($P4MASTER_ID) does not start with 'master'." fi fi fi # TODO: # - Ensure checkpoints dir contains a checkpoint or two. # - service password verified as set for a replica (value doesn't matter, it just needs to be set). # - Add flag to check less-critical SDP configurables, and generate # warnings (rather than errors) if they are not set as expected, # using SDP configure_new_servers.sh script as a guide. if [[ "$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 | |
---|---|---|---|---|---|
#20 | 30915 | C. Thomas Tyler |
Released SDP 2024.1.30913 (2024/11/20). Copy Up using 'p4 copy -r -b perforce_software-sdp-dev'. |
||
#19 | 30297 | C. Thomas Tyler |
Released SDP 2023.2.30295 (2024/05/08). Copy Up using 'p4 copy -r -b perforce_software-sdp-dev'. |
||
#18 | 30043 | C. Thomas Tyler |
Released SDP 2023.2.30041 (2023/12/22). Copy Up using 'p4 copy -r -b perforce_software-sdp-dev'. |
||
#17 | 29143 | C. Thomas Tyler |
Released SDP 2022.1.29141 (2022/10/29). Copy Up using 'p4 copy -r -b perforce_software-sdp-dev'. |
||
#16 | 28858 | C. Thomas Tyler |
Released SDP 2022.1.28855 (2022/05/27). Copy Up using 'p4 copy -r -b perforce_software-sdp-dev'. |
||
#15 | 28651 | C. Thomas Tyler |
Released SDP 2021.2.28649 (2022/03/03). Copy Up using 'p4 copy -r -b perforce_software-sdp-dev'. |
||
#14 | 28259 | C. Thomas Tyler |
Released SDP 2021.1.28253 (2021/11/13). Copy Up using 'p4 copy -r -b perforce_software-sdp-dev'. |
||
#13 | 28240 | C. Thomas Tyler |
Released SDP 2021.1.28238 (2021/11/12). Copy Up using 'p4 copy -r -b perforce_software-sdp-dev'. |
||
#12 | 27921 | C. Thomas Tyler |
Released SDP 2020.1.27919 (2021/07/19). Copy Up using 'p4 copy -r -b perforce_software-sdp-dev'. |
||
#11 | 27761 | C. Thomas Tyler |
Released SDP 2020.1.27759 (2021/05/07). Copy Up using 'p4 copy -r -b perforce_software-sdp-dev'. |
||
#10 | 27416 | C. Thomas Tyler |
Released SDP 2020.1.27414 (2021/02/07). Copy Up using 'p4 copy -r -b perforce_software-sdp-dev'. |
||
#9 | 27407 | C. Thomas Tyler |
Released SDP 2020.1.27403 (2021/02/06). Copy Up using 'p4 copy -r -b perforce_software-sdp-dev'. |
||
#8 | 27400 | C. Thomas Tyler |
Released SDP 2020.1.27398 (2021/02/06). Copy Up using 'p4 copy -r -b perforce_software-sdp-dev'. |
||
#7 | 27354 | C. Thomas Tyler |
Released SDP 2020.1.27351 (2021/01/31). Copy Up using 'p4 copy -r -b perforce_software-sdp-dev'. |
||
#6 | 27331 | C. Thomas Tyler |
Released SDP 2020.1.27325 (2021/01/29). Copy Up using 'p4 copy -r -b perforce_software-sdp-dev'. |
||
#5 | 26470 | C. Thomas Tyler |
Released SDP 2019.3.26468 (2020/04/10). Copy Up using 'p4 copy -r -b perforce_software-sdp-dev'. |
||
#4 | 26403 | C. Thomas Tyler |
Released SDP 2019.3.26400 (2020/03/28). Copy Up using 'p4 copy -r -b perforce_software-sdp-dev'. |
||
#3 | 26161 | C. Thomas Tyler |
Released SDP 2019.3.26159 (2019/11/06). Copy Up using 'p4 copy -r -b perforce_software-sdp-dev'. |
||
#2 | 25380 | C. Thomas Tyler |
Released SDP 2019.1.25374 (2019/03/21). Copy Up using 'p4 copy -r -b perforce_software-sdp-dev'. |
||
#1 | 25245 | C. Thomas Tyler |
Released SDP 2019.1.25238 (2019/03/02). Copy Up using 'p4 copy -r -b perforce_software-sdp-dev'. |
||
//guest/perforce_software/sdp/dev/Server/Unix/p4/common/bin/verify_sdp.sh | |||||
#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! |