if [ -z "${LOG_TITLE}" ]; then LOG_TITLE='' fi if [ -z "${LOG_LEVEL}" ]; then LOG_LEVEL="INFO" fi
debug() { if [[ "${LOG_LEVEL}" == "DEBUG" ]]; then local log_title if [ -n "${LOG_TITLE}" ]; then log_title="(${LOG_TITLE})" else log_title='' fi echo -e "${GREEN}[DEBUG]${log_title}${NO_COLOR}$1" fi }
info() { if [[ "${LOG_LEVEL}" == "DEBUG" ]] ||\ [[ "${LOG_LEVEL}" == "INFO" ]]; then local log_title if [ -n "${LOG_TITLE}" ]; then log_title="(${LOG_TITLE})" else log_title='' fi echo -e "${CYAN}[INFO] ${log_title}${NO_COLOR}$1" fi }
warn() { if [[ "${LOG_LEVEL}" == "DEBUG" ]] ||\ [[ "${LOG_LEVEL}" == "INFO" ]] ||\ [[ "${LOG_LEVEL}" == "WARN" ]]; then local log_title if [ -n "${LOG_TITLE}" ]; then log_title="(${LOG_TITLE})" else log_title='' fi echo -e "${YELLOW}[WARN] ${log_title}${NO_COLOR}$1" fi }
error() { if [[ "${LOG_LEVEL}" == "DEBUG" ]] ||\ [[ "${LOG_LEVEL}" == "INFO" ]] ||\ [[ "${LOG_LEVEL}" == "WARN" ]] ||\ [[ "${LOG_LEVEL}" == "ERROR" ]]; then local log_title if [ -n "${LOG_TITLE}" ]; then log_title="(${LOG_TITLE})" else log_title='' fi echo -e "${RED}[ERROR]${log_title}${NO_COLOR}$1" fi }
local all_found=true for ((i=0; i<${#targets[@]}; i++)); do local target=${targets[$i]} if [ "$(which $target)" = "" ]; then all_found=false error "Not found: $target" fi done
if [ "$all_found" = "false" ]; then msg="Please install missing dependencies: ${targets[@]}." info "$msg" exit 2 fi
msg="Required dependencies '${targets[@]}' are installed." info "$msg" }
cleanup() { info "Cleaning up longhorn-environment-check pods..." kubectl delete -f $TEMP_DIR/environment_check.yaml > /dev/null rm -rf $TEMP_DIR info "Cleanup completed." }
wait_ds_ready() { whiletrue; do local ds=$(kubectl get ds/longhorn-environment-check -o json) local numberReady=$(echo$ds | jq .status.numberReady) local desiredNumberScheduled=$(echo$ds | jq .status.desiredNumberScheduled)
if [ "$desiredNumberScheduled" = "$numberReady" ] && [ "$desiredNumberScheduled" != "0" ]; then info "All longhorn-environment-check pods are ready ($numberReady/$desiredNumberScheduled)." return fi
info "Waiting for longhorn-environment-check pods to become ready ($numberReady/$desiredNumberScheduled)..." sleep 3 done }
check_mount_propagation() { local allSupported=true local pods=$(kubectl -l app=longhorn-environment-check get po -o json)
local ds=$(kubectl get ds/longhorn-environment-check -o json) local desiredNumberScheduled=$(echo$ds | jq .status.desiredNumberScheduled)
for ((i=0; i<desiredNumberScheduled; i++)); do local pod=$(echo$pods | jq .items[$i]) local nodeName=$(echo$pod | jq -r .spec.nodeName) local mountPropagation=$(echo$pod | jq -r '.spec.containers[0].volumeMounts[] | select(.name=="mountpoint") | .mountPropagation')
if [ "$mountPropagation" != "Bidirectional" ]; then allSupported=false error "node $nodeName: MountPropagation is disabled" fi done
if [ "$allSupported" != "true" ]; then error "MountPropagation is disabled on at least one node. As a result, CSI driver and Base image cannot be supported" exit 1 else info "MountPropagation is enabled" fi }
check_hostname_uniqueness() { hostnames=$(kubectl get nodes -o jsonpath='{.items[*].status.addresses[?(@.type=="Hostname")].address}')
if [ $? -ne 0 ]; then error "kubectl get nodes failed - check KUBECONFIG setup" exit 1 fi
if [[ ! ${hostnames[@]} ]]; then error "kubectl get nodes returned empty list - check KUBECONFIG setup" exit 1 fi
deduplicate_hostnames=() num_nodes=0 for hostname in${hostnames}; do num_nodes=$((num_nodes+1)) if ! echo"${deduplicate_hostnames[@]}" | grep -q "\<${hostname}\>"; then deduplicate_hostnames+=("${hostname}") fi done
if [ "${#deduplicate_hostnames[@]}" != "${num_nodes}" ]; then error "Nodes do not have unique hostnames." exit 2 fi
info "All nodes have unique hostnames." }
check_nodes() { local name=$1 local callback=$2 shift shift
info "Checking $name..."
local all_passed=true
local pods=$(kubectl get pods -o name -l app=longhorn-environment-check) for pod in${pods}; do eval"${callback}${pod}$@" if [ $? -ne 0 ]; then all_passed=false fi done
if [ "$all_passed" = "false" ]; then return 1 fi }
check_kernel_release() { local pod=$1 local node=$(kubectl get ${pod} --no-headers -o=custom-columns=:.spec.nodeName)
recommended_kernel_release="5.8"
local kernel=$(detect_node_kernel_release ${pod})
if verlt "$kernel""$recommended_kernel_release" ; then warn "Node $node has outdated kernel release: $kernel. Recommending kernel release >= $recommended_kernel_release" return 1 fi
local broken_kernel=("5.15.0-94""6.5.6") local fixed_kernel=("5.15.0-100""6.5.7")
for i in${!broken_kernel[@]}; do if kernel_in_range "$kernel""${broken_kernel[$i]}""${fixed_kernel[$i]}" ; then warn "Node $node has a kernel version $kernel known to have a breakage that affects Longhorn. See description and solution at https://longhorn.io/kb/troubleshooting-rwx-volume-fails-to-attached-caused-by-protocol-not-supported" return 1 fi done
}
check_iscsid() { local pod=$1
kubectl exec${pod} -- nsenter --mount=/proc/1/ns/mnt -- bash -c "systemctl status --no-pager iscsid.service" > /dev/null 2>&1 if [ $? -ne 0 ]; then kubectl exec${pod} -- nsenter --mount=/proc/1/ns/mnt -- bash -c "systemctl status --no-pager iscsid.socket" > /dev/null 2>&1 if [ $? -ne 0 ]; then node=$(kubectl get ${pod} --no-headers -o=custom-columns=:.spec.nodeName) error "Neither iscsid.service nor iscsid.socket is running on ${node}" return 1 fi fi }
check_multipathd() { local pod=$1
kubectl exec$pod -- nsenter --mount=/proc/1/ns/mnt -- bash -c "systemctl status --no-pager multipathd.service" > /dev/null 2>&1 if [ $? = 0 ]; then node=$(kubectl get ${pod} --no-headers -o=custom-columns=:.spec.nodeName) warn "multipathd is running on ${node} known to have a breakage that affects Longhorn. See description and solution at https://longhorn.io/kb/troubleshooting-volume-with-multipath" return 1 fi }
check_packages() { local pod=$1
OS=$(detect_node_os ${pod}) if [ x"$OS" = x"" ]; then error "Failed to detect OS on node ${node}" return 1 fi
set_packages_and_check_cmd
for ((i=0; i<${#PACKAGES[@]}; i++)); do check_package ${PACKAGES[$i]} if [ $? -ne 0 ]; then return 1 fi done }
check_package() { local package=$1
kubectl exec$pod -- nsenter --mount=/proc/1/ns/mnt -- timeout 30 bash -c "$CHECK_CMD$package" > /dev/null 2>&1 if [ $? -ne 0 ]; then node=$(kubectl get ${pod} --no-headers -o=custom-columns=:.spec.nodeName) error "$package is not found in $node." return 1 fi }
check_nfs_client() { local pod=$1 local node=$(kubectl get ${pod} --no-headers -o=custom-columns=:.spec.nodeName)
local options=("CONFIG_NFS_V4_2""CONFIG_NFS_V4_1""CONFIG_NFS_V4")
local kernel=$(detect_node_kernel_release ${pod}) if [ "x${kernel}" = "x" ]; then warn "Failed to check NFS client installation, because unable to detect kernel release on node ${node}" return 1 fi
for option in"${options[@]}"; do kubectl exec${pod} -- nsenter --mount=/proc/1/ns/mnt -- bash -c "[ -f /boot/config-${kernel} ]" > /dev/null 2>&1 if [ $? -ne 0 ]; then warn "Failed to check $option on node ${node}, because /boot/config-${kernel} does not exist on node ${node}" continue fi
check_kernel_module ${pod}${option} nfs if [ $? = 0 ]; then return 0 fi done
error "NFS clients ${options[*]} not found. At least one should be enabled" return 1 }
check_kernel_module() { local pod=$1 local option=$2 local module=$3
local kernel=$(detect_node_kernel_release ${pod}) if [ "x${kernel}" = "x" ]; then warn "Failed to check kernel config option ${option}, because unable to detect kernel release on node ${node}" return 1 fi
kubectl exec${pod} -- nsenter --mount=/proc/1/ns/mnt -- bash -c "[ -e /boot/config-${kernel} ]" > /dev/null 2>&1 if [ $? -ne 0 ]; then warn "Failed to check kernel config option ${option}, because /boot/config-${kernel} does not exist on node ${node}" return 1 fi
value=$(kubectl exec${pod} -- nsenter --mount=/proc/1/ns/mnt -- bash -c "grep "^$option=" /boot/config-${kernel} | cut -d= -f2") if [ -z "${value}" ]; then error "Failed to find kernel config $option on node ${node}" return 1 elif [ "${value}" = "m" ]; then kubectl exec${pod} -- nsenter --mount=/proc/1/ns/mnt -- bash -c "lsmod | grep ${module}" > /dev/null 2>&1 if [ $? -ne 0 ]; then node=$(kubectl get ${pod} --no-headers -o=custom-columns=:.spec.nodeName) error "kernel module ${module} is not enabled on ${node}" return 1 fi elif [ "${value}" = "y" ]; then return 0 else warn "Unknown value for $option: $value" return 1 fi }
check_hugepage() { local pod=$1 local expected_nr_hugepages=$2
nr_hugepages=$(kubectl exec${pod} -- nsenter --mount=/proc/1/ns/mnt -- bash -c 'cat /proc/sys/vm/nr_hugepages') if [ $? -ne 0 ]; then error "Failed to check hugepage size on node ${node}" return 1 fi
if [ $nr_hugepages -lt $expected_nr_hugepages ]; then error "Hugepage size is not enough on node ${node}. Expected: ${expected_nr_hugepages}, Actual: ${nr_hugepages}" return 1 fi }
functioncheck_sse42_support() { local pod=$1
node=$(kubectl get ${pod} --no-headers -o=custom-columns=:.spec.nodeName)
machine=$(kubectl exec$pod -- nsenter --mount=/proc/1/ns/mnt -- bash -c 'uname -m' 2>/dev/null) if [ $? -ne 0 ]; then error "Failed to check machine on node ${node}" return 1 fi
if [ "$machine" = "x86_64" ]; then sse42_support=$(kubectl exec$pod -- nsenter --mount=/proc/1/ns/mnt -- bash -c 'grep -o sse4_2 /proc/cpuinfo | wc -l' 2>/dev/null) if [ $? -ne 0 ]; then error "Failed to check SSE4.2 instruction set on node ${node}" return 1 fi
if [ "$sse42_support" -ge 1 ]; then return 0 fi
error "CPU does not support SSE4.2" return 1 else warn "Skip SSE4.2 instruction set check on node ${node} because it is not x86_64" fi }
functionshow_help() { cat <<EOF Usage: $0 [OPTIONS] Options: -s, --enable-spdk Enable checking SPDK prerequisites -p, --expected-nr-hugepages Expected number of 2 MiB hugepages for SPDK. Default: 1024 -h, --help Show this help message and exit EOF exit 0 }
http://mirrors.bfsu.edu.cn/centos/7.9.2009/updates/x86_64/Packages/iscsi-initiator-utils-iscsiuio-6.2.0.874-22.el7_9.x86_64.rpm: [Errno 14] HTTP Error 404 - Not Found--:-- ETA Trying other mirror. iscsi-initiator-utils-6.2.0.87 FAILED http://mirrors.huaweicloud.com/centos/7.9.2009/updates/x86_64/Packages/iscsi-initiator-utils-6.2.0.874-22.el7_9.x86_64.rpm: [Errno 14] HTTP Error 404 - Not Found --:--:-- ETA Trying other mirror. iscsi-initiator-utils-iscsiuio FAILED http://mirrors.163.com/centos/7.9.2009/updates/x86_64/Packages/iscsi-initiator-utils-iscsiuio-6.2.0.874-22.el7_9.x86_64.rpm: [Errno 14] HTTP Error 404 - Not Found --:--:-- ETA Trying other mirror.
Error downloading packages: iscsi-initiator-utils-6.2.0.874-22.el7_9.x86_64: [Errno 256] No more mirrors to try. iscsi-initiator-utils-iscsiuio-6.2.0.874-22.el7_9.x86_64: [Errno 256] No more mirrors to try.
# CentOS-Base.repo # # The mirror system uses the connecting IP address of the client and the # update status of each mirror to pick mirrors that are updated to and # geographically close to the client. You should use this for CentOS updates # unless you are manually picking other mirrors. # # If the mirrorlist= does not work for you, as a fall back you can try the # remarked out baseurl= line instead. # #
