diff --git a/.github/workflows/AlmaLinux.yml b/.github/workflows/AlmaLinux.yml deleted file mode 100644 index dd06dd2..0000000 --- a/.github/workflows/AlmaLinux.yml +++ /dev/null @@ -1,64 +0,0 @@ -# See: https://docs.github.com/en/actions/using-workflows/workflow-syntax-for-github-actions - -# Workflow name -name: AlmaLinux 8.5 RPM build - -# Run on tag push -on: - push: - tags: - - '**' - -jobs: - - # - # Build on AlmaLinux 8.5 using go-toolset - # - AlmaLinux-RPM-build: - runs-on: ubuntu-latest - # See: https://hub.docker.com/_/almalinux - container: almalinux:8.5 - steps: - - # Use dnf to install development packages - - name: Install development packages - run: dnf --assumeyes group install "Development Tools" "RPM Development Tools" - - # Checkout git repository and submodules - # fetch-depth must be 0 to use git describe - # See: https://github.com/marketplace/actions/checkout - - name: Checkout - uses: actions/checkout@v2 - with: - submodules: recursive - fetch-depth: 0 - - # Use dnf to install build dependencies - - name: Install build dependencies - run: dnf --assumeyes builddep scripts/cc-metric-collector.spec - - - name: RPM build MetricCollector - id: rpmbuild - run: make RPM - - # See: https://github.com/actions/upload-artifact - - name: Save RPM as artifact - uses: actions/upload-artifact@v2 - with: - name: cc-metric-collector RPM for AlmaLinux 8.5 - path: ${{ steps.rpmbuild.outputs.RPM }} - - name: Save SRPM as artifact - uses: actions/upload-artifact@v2 - with: - name: cc-metric-collector SRPM for AlmaLinux 8.5 - path: ${{ steps.rpmbuild.outputs.SRPM }} - - # See: https://github.com/softprops/action-gh-release - - name: Release - uses: softprops/action-gh-release@v1 - if: startsWith(github.ref, 'refs/tags/') - with: - name: cc-metric-collector-${{github.ref_name}} - files: | - ${{ steps.rpmbuild.outputs.RPM }} - ${{ steps.rpmbuild.outputs.SRPM }} diff --git a/.github/workflows/RedHatUniversalBaseImage.yml b/.github/workflows/RedHatUniversalBaseImage.yml deleted file mode 100644 index 205a133..0000000 --- a/.github/workflows/RedHatUniversalBaseImage.yml +++ /dev/null @@ -1,64 +0,0 @@ -# See: https://docs.github.com/en/actions/using-workflows/workflow-syntax-for-github-actions - -# Workflow name -name: Red Hat Universal Base Image 8 RPM build - -# Run on tag push -on: - push: - tags: - - '**' - -jobs: - - # - # Build on UBI 8 using go-toolset - # - UBI-8-RPM-build: - runs-on: ubuntu-latest - # See: https://catalog.redhat.com/software/containers/ubi8/ubi/5c359854d70cc534b3a3784e?container-tabs=gti - container: registry.access.redhat.com/ubi8/ubi:8.5-226.1645809065 - steps: - - # Use dnf to install development packages - - name: Install development packages - run: dnf --assumeyes --disableplugin=subscription-manager install rpm-build go-srpm-macros rpm-build-libs rpm-libs gcc make python38 git - - # Checkout git repository and submodules - # fetch-depth must be 0 to use git describe - # See: https://github.com/marketplace/actions/checkout - - name: Checkout - uses: actions/checkout@v2 - with: - submodules: recursive - fetch-depth: 0 - - # Use dnf to install build dependencies - - name: Install build dependencies - run: dnf --assumeyes --disableplugin=subscription-manager builddep scripts/cc-metric-collector.spec - - - name: RPM build MetricCollector - id: rpmbuild - run: make RPM - - # See: https://github.com/actions/upload-artifact - - name: Save RPM as artifact - uses: actions/upload-artifact@v2 - with: - name: cc-metric-collector RPM for UBI 8 - path: ${{ steps.rpmbuild.outputs.RPM }} - - name: Save SRPM as artifact - uses: actions/upload-artifact@v2 - with: - name: cc-metric-collector SRPM for UBI 8 - path: ${{ steps.rpmbuild.outputs.SRPM }} - - # See: https://github.com/softprops/action-gh-release - - name: Release - uses: softprops/action-gh-release@v1 - if: startsWith(github.ref, 'refs/tags/') - with: - name: cc-metric-collector-${{github.ref_name}} - files: | - ${{ steps.rpmbuild.outputs.RPM }} - ${{ steps.rpmbuild.outputs.SRPM }} diff --git a/.github/workflows/Release.yml b/.github/workflows/Release.yml new file mode 100644 index 0000000..9ce5615 --- /dev/null +++ b/.github/workflows/Release.yml @@ -0,0 +1,184 @@ +# See: https://docs.github.com/en/actions/using-workflows/workflow-syntax-for-github-actions + +# Workflow name +name: Release + +# Run on tag push +on: + push: + tags: + - '**' + +jobs: + + # + # Build on AlmaLinux 8.5 using go-toolset + # + AlmaLinux-RPM-build: + runs-on: ubuntu-latest + # See: https://hub.docker.com/_/almalinux + container: almalinux:8.5 + # The job outputs link to the outputs of the 'rpmrename' step + # Only job outputs can be used in child jobs + outputs: + rpm : ${{steps.rpmrename.outputs.RPM}} + srpm : ${{steps.rpmrename.outputs.SRPM}} + steps: + + # Use dnf to install development packages + - name: Install development packages + run: dnf --assumeyes group install "Development Tools" "RPM Development Tools" + + # Checkout git repository and submodules + # fetch-depth must be 0 to use git describe + # See: https://github.com/marketplace/actions/checkout + - name: Checkout + uses: actions/checkout@v2 + with: + submodules: recursive + fetch-depth: 0 + + # Use dnf to install build dependencies + - name: Install build dependencies + run: dnf --assumeyes builddep scripts/cc-metric-collector.spec + + - name: RPM build MetricCollector + id: rpmbuild + run: make RPM + + # AlmaLinux 8.5 is a derivate of RedHat Enterprise Linux 8 (UBI8), + # so the created RPM both contain the substring 'el8' in the RPM file names + # This step replaces the substring 'el8' to 'alma85'. It uses the move operation + # because it is unclear whether the default AlmaLinux 8.5 container contains the + # 'rename' command. This way we also get the new names for output. + - name: Rename RPMs (s/el8/alma85/) + id: rpmrename + run: | + OLD_RPM="${{steps.rpmbuild.outputs.RPM}}" + OLD_SRPM="${{steps.rpmbuild.outputs.SRPM}}" + NEW_RPM="${OLD_RPM/el8/alma85}" + NEW_SRPM=${OLD_SRPM/el8/alma85} + mv "${OLD_RPM}" "${NEW_RPM}" + mv "${OLD_SRPM}" "${NEW_SRPM}" + echo "::set-output name=SRPM::${NEW_SRPM}" + echo "::set-output name=RPM::${NEW_RPM}" + + # See: https://github.com/actions/upload-artifact + - name: Save RPM as artifact + uses: actions/upload-artifact@v2 + with: + name: cc-metric-collector RPM for AlmaLinux 8.5 + path: ${{ steps.rpmrename.outputs.RPM }} + - name: Save SRPM as artifact + uses: actions/upload-artifact@v2 + with: + name: cc-metric-collector SRPM for AlmaLinux 8.5 + path: ${{ steps.rpmrename.outputs.SRPM }} + + # + # Build on UBI 8 using go-toolset + # + UBI-8-RPM-build: + runs-on: ubuntu-latest + # See: https://catalog.redhat.com/software/containers/ubi8/ubi/5c359854d70cc534b3a3784e?container-tabs=gti + container: registry.access.redhat.com/ubi8/ubi:8.5-226.1645809065 + # The job outputs link to the outputs of the 'rpmbuild' step + outputs: + rpm : ${{steps.rpmbuild.outputs.RPM}} + srpm : ${{steps.rpmbuild.outputs.SRPM}} + steps: + + # Use dnf to install development packages + - name: Install development packages + run: dnf --assumeyes --disableplugin=subscription-manager install rpm-build go-srpm-macros rpm-build-libs rpm-libs gcc make python38 git + + # Checkout git repository and submodules + # fetch-depth must be 0 to use git describe + # See: https://github.com/marketplace/actions/checkout + - name: Checkout + uses: actions/checkout@v2 + with: + submodules: recursive + fetch-depth: 0 + + # Use dnf to install build dependencies + - name: Install build dependencies + run: dnf --assumeyes --disableplugin=subscription-manager builddep scripts/cc-metric-collector.spec + + - name: RPM build MetricCollector + id: rpmbuild + run: make RPM + + # See: https://github.com/actions/upload-artifact + - name: Save RPM as artifact + uses: actions/upload-artifact@v2 + with: + name: cc-metric-collector RPM for UBI 8 + path: ${{ steps.rpmbuild.outputs.RPM }} + - name: Save SRPM as artifact + uses: actions/upload-artifact@v2 + with: + name: cc-metric-collector SRPM for UBI 8 + path: ${{ steps.rpmbuild.outputs.SRPM }} + + # + # Create release with fresh RPMs + # + Release: + runs-on: ubuntu-latest + # We need the RPMs, so add dependency + needs: [AlmaLinux-RPM-build, UBI-8-RPM-build] + + steps: + # See: https://github.com/actions/download-artifact + - name: Download AlmaLinux 8.5 RPM + uses: actions/download-artifact@v2 + with: + name: cc-metric-collector RPM for AlmaLinux 8.5 + - name: Download AlmaLinux 8.5 SRPM + uses: actions/download-artifact@v2 + with: + name: cc-metric-collector SRPM for AlmaLinux 8.5 + + - name: Download UBI 8 RPM + uses: actions/download-artifact@v2 + with: + name: cc-metric-collector RPM for UBI 8 + - name: Download UBI 8 SRPM + uses: actions/download-artifact@v2 + with: + name: cc-metric-collector SRPM for UBI 8 + + # The download actions do not publish the name of the downloaded file, + # so we re-use the job outputs of the parent jobs. The files are all + # downloaded to the current folder. + # The gh-release action afterwards does not accept file lists but all + # files have to be listed at 'files'. The step creates one output per + # RPM package (2 per distro) + - name: Set RPM variables + id: files + run: | + ALMA_85_RPM=$(basename "${{ needs.AlmaLinux-RPM-build.outputs.rpm}}") + ALMA_85_SRPM=$(basename "${{ needs.AlmaLinux-RPM-build.outputs.srpm}}") + UBI_8_RPM=$(basename "${{ needs.UBI-8-RPM-build.outputs.rpm}}") + UBI_8_SRPM=$(basename "${{ needs.UBI-8-RPM-build.outputs.srpm}}") + echo "ALMA_85_RPM::${ALMA_85_RPM}" + echo "ALMA_85_SRPM::${ALMA_85_SRPM}" + echo "UBI_8_RPM::${UBI_8_RPM}" + echo "UBI_8_SRPM::${UBI_8_SRPM}" + echo "::set-output name=ALMA_85_RPM::${ALMA_85_RPM}" + echo "::set-output name=ALMA_85_SRPM::${ALMA_85_SRPM}" + echo "::set-output name=UBI_8_RPM::${UBI_8_RPM}" + echo "::set-output name=UBI_8_SRPM::${UBI_8_SRPM}" + + # See: https://github.com/softprops/action-gh-release + - name: Release + uses: softprops/action-gh-release@v1 + if: startsWith(github.ref, 'refs/tags/') + with: + name: cc-metric-collector-${{github.ref_name}} + files: | + ${{ steps.files.outputs.ALMA_85_RPM }} + ${{ steps.files.outputs.ALMA_85_SRPM }} + ${{ steps.files.outputs.UBI_8_RPM }} + ${{ steps.files.outputs.UBI_8_SRPM }} \ No newline at end of file diff --git a/Makefile b/Makefile index d747899..0a7ad04 100644 --- a/Makefile +++ b/Makefile @@ -15,6 +15,8 @@ COMPONENT_DIRS := collectors \ internal/ccTopology \ internal/multiChanTicker +BINDIR = bin + .PHONY: all all: $(APP) @@ -24,15 +26,27 @@ $(APP): $(GOSRC) go get go build -o $(APP) $(GOSRC_APP) +install: $(APP) + @WORKSPACE=$(PREFIX) + @if [ -z "$${WORKSPACE}" ]; then exit 1; fi + @mkdir --parents --verbose $${WORKSPACE}/usr/$(BINDIR) + @install -Dpm 755 $(APP) $${WORKSPACE}/usr/$(BINDIR)/$(APP) + @mkdir --parents --verbose $${WORKSPACE}/etc/cc-metric-collector $${WORKSPACE}/etc/default $${WORKSPACE}/etc/systemd/system $${WORKSPACE}/etc/init.d + @install -Dpm 600 config.json $${WORKSPACE}/etc/cc-metric-collector/cc-metric-collector.json + @sed -i -e s+"\"./"+"\"/etc/cc-metric-collector/"+g $${WORKSPACE}/etc/cc-metric-collector/cc-metric-collector.json + @install -Dpm 600 sinks.json $${WORKSPACE}/etc/cc-metric-collector/sinks.json + @install -Dpm 600 collectors.json $${WORKSPACE}/etc/cc-metric-collector/collectors.json + @install -Dpm 600 router.json $${WORKSPACE}/etc/cc-metric-collector/router.json + @install -Dpm 600 receivers.json $${WORKSPACE}/etc/cc-metric-collector/receivers.json + @install -Dpm 600 scripts/cc-metric-collector.config $${WORKSPACE}/etc/default/cc-metric-collector + @install -Dpm 644 scripts/cc-metric-collector.service $${WORKSPACE}/etc/systemd/system/cc-metric-collector.service + @install -Dpm 644 scripts/cc-metric-collector.init $${WORKSPACE}/etc/init.d/cc-metric-collector + + .PHONY: clean .ONESHELL: clean: - @for COMP in $(COMPONENT_DIRS) - do - if [[ -e $$COMP/Makefile ]]; then - make -C $$COMP clean - fi - done + @for COMP in $(COMPONENT_DIRS); do if [ -e $$COMP/Makefile ]; then make -C $$COMP clean; fi; done rm -f $(APP) .PHONY: fmt @@ -69,7 +83,7 @@ RPM: scripts/cc-metric-collector.spec @COMMITISH="HEAD" @VERS=$$(git describe --tags $${COMMITISH}) @VERS=$${VERS#v} - @VERS=$${VERS//-/_} + @VERS=$$(echo $$VERS | sed -e s+'-'+'_'+g) @eval $$(rpmspec --query --queryformat "NAME='%{name}' VERSION='%{version}' RELEASE='%{release}' NVR='%{NVR}' NVRA='%{NVRA}'" --define="VERS $${VERS}" "$${SPECFILE}") @PREFIX="$${NAME}-$${VERSION}" @FORMAT="tar.gz" @@ -86,3 +100,28 @@ RPM: scripts/cc-metric-collector.spec @ echo "::set-output name=SRPM::$${SRPMFILE}" @ echo "::set-output name=RPM::$${RPMFILE}" @fi + +.PHONY: DEB +DEB: scripts/cc-metric-collector.deb.control $(APP) + @BASEDIR=$${PWD} + @WORKSPACE=$${PWD}/.dpkgbuild + @DEBIANDIR=$${WORKSPACE}/debian + @DEBIANBINDIR=$${WORKSPACE}/DEBIAN + @mkdir --parents --verbose $$WORKSPACE $$DEBIANBINDIR + #@mkdir --parents --verbose $$DEBIANDIR + @CONTROLFILE="$${BASEDIR}/scripts/cc-metric-collector.deb.control" + @COMMITISH="HEAD" + @VERS=$$(git describe --tags --abbrev=0 $${COMMITISH}) + @VERS=$${VERS#v} + @VERS=$$(echo $$VERS | sed -e s+'-'+'_'+g) + @ARCH=$$(uname -m) + @ARCH=$$(echo $$ARCH | sed -e s+'_'+'-'+g) + @PREFIX="$${NAME}-$${VERSION}_$${ARCH}" + @SIZE_BYTES=$$(du -bcs --exclude=.dpkgbuild "$$WORKSPACE"/ | awk '{print $$1}' | head -1 | sed -e 's/^0\+//') + @SIZE="$$(awk -v size="$$SIZE_BYTES" 'BEGIN {print (size/1024)+1}' | awk '{print int($$0)}')" + #@sed -e s+"{VERSION}"+"$$VERS"+g -e s+"{INSTALLED_SIZE}"+"$$SIZE"+g -e s+"{ARCH}"+"$$ARCH"+g $$CONTROLFILE > $${DEBIANDIR}/control + @sed -e s+"{VERSION}"+"$$VERS"+g -e s+"{INSTALLED_SIZE}"+"$$SIZE"+g -e s+"{ARCH}"+"$$ARCH"+g $$CONTROLFILE > $${DEBIANBINDIR}/control + @make PREFIX=$${WORKSPACE} install + @DEB_FILE="cc-metric-collector_$${VERS}_$${ARCH}.deb" + @dpkg-deb -b $${WORKSPACE} "$$DEB_FILE" + @rm -r "$${WORKSPACE}" diff --git a/collectors.json b/collectors.json index 27ef822..be3eea7 100644 --- a/collectors.json +++ b/collectors.json @@ -12,6 +12,12 @@ "proc_total" ] }, + "netstat": { + "include_devices": [ + "enp5s0" + ], + "send_derived_values": true + }, "numastats": {}, "nvidia": {}, "tempstat": { diff --git a/collectors/gpfsMetric.go b/collectors/gpfsMetric.go index 453704c..26fc723 100644 --- a/collectors/gpfsMetric.go +++ b/collectors/gpfsMetric.go @@ -17,7 +17,12 @@ import ( lp "github.com/ClusterCockpit/cc-metric-collector/internal/ccMetric" ) -const DEFAULT_GPFS_CMD = `mmpmon` +const DEFAULT_GPFS_CMD = "mmpmon" + +type GpfsCollectorLastState struct { + bytesRead int64 + bytesWritten int64 +} type GpfsCollector struct { metricCollector @@ -25,8 +30,11 @@ type GpfsCollector struct { config struct { Mmpmon string `json:"mmpmon_path,omitempty"` ExcludeFilesystem []string `json:"exclude_filesystem,omitempty"` + SendBandwidths bool `json:"send_bandwidths"` } - skipFS map[string]struct{} + skipFS map[string]struct{} + lastTimestamp time.Time // Store time stamp of last tick to derive bandwidths + lastState map[string]GpfsCollectorLastState } func (m *GpfsCollector) Init(config json.RawMessage) error { @@ -40,7 +48,7 @@ func (m *GpfsCollector) Init(config json.RawMessage) error { m.setup() // Set default mmpmon binary - m.config.Mmpmon = string(DEFAULT_GPFS_CMD) + m.config.Mmpmon = DEFAULT_GPFS_CMD // Read JSON configuration if len(config) > 0 { @@ -89,6 +97,13 @@ func (m *GpfsCollector) Read(interval time.Duration, output chan lp.CCMetric) { return } + // Current time stamp + now := time.Now() + // time difference to last time stamp + timeDiff := now.Sub(m.lastTimestamp).Seconds() + // Save current timestamp + m.lastTimestamp = now + // mmpmon: // -p: generate output that can be parsed // -s: suppress the prompt on input @@ -148,6 +163,12 @@ func (m *GpfsCollector) Read(interval time.Duration, output chan lp.CCMetric) { } m.tags["filesystem"] = filesystem + if _, ok := m.lastState[filesystem]; !ok { + m.lastState[filesystem] = GpfsCollectorLastState{ + bytesRead: -1, + bytesWritten: -1, + } + } // return code rc, err := strconv.Atoi(key_value["_rc_"]) @@ -191,6 +212,14 @@ func (m *GpfsCollector) Read(interval time.Duration, output chan lp.CCMetric) { if y, err := lp.New("gpfs_bytes_read", m.tags, m.meta, map[string]interface{}{"value": bytesRead}, timestamp); err == nil { output <- y } + if m.config.SendBandwidths { + if lastBytesRead := m.lastState[filesystem].bytesRead; lastBytesRead >= 0 { + bwRead := float64(bytesRead-lastBytesRead) / timeDiff + if y, err := lp.New("gpfs_bw_read", m.tags, m.meta, map[string]interface{}{"value": bwRead}, timestamp); err == nil { + output <- y + } + } + } // bytes written bytesWritten, err := strconv.ParseInt(key_value["_bw_"], 10, 64) @@ -203,6 +232,21 @@ func (m *GpfsCollector) Read(interval time.Duration, output chan lp.CCMetric) { if y, err := lp.New("gpfs_bytes_written", m.tags, m.meta, map[string]interface{}{"value": bytesWritten}, timestamp); err == nil { output <- y } + if m.config.SendBandwidths { + if lastBytesWritten := m.lastState[filesystem].bytesRead; lastBytesWritten >= 0 { + bwWrite := float64(bytesWritten-lastBytesWritten) / timeDiff + if y, err := lp.New("gpfs_bw_write", m.tags, m.meta, map[string]interface{}{"value": bwWrite}, timestamp); err == nil { + output <- y + } + } + } + + if m.config.SendBandwidths { + m.lastState[filesystem] = GpfsCollectorLastState{ + bytesRead: bytesRead, + bytesWritten: bytesWritten, + } + } // number of opens numOpens, err := strconv.ParseInt(key_value["_oc_"], 10, 64) diff --git a/collectors/gpfsMetric.md b/collectors/gpfsMetric.md index 4f2c897..30a5a40 100644 --- a/collectors/gpfsMetric.md +++ b/collectors/gpfsMetric.md @@ -5,7 +5,8 @@ "mmpmon_path": "/path/to/mmpmon", "exclude_filesystem": [ "fs1" - ] + ], + "send_bandwidths" : true } ``` @@ -18,13 +19,16 @@ in the configuration. The path to the `mmpmon` command can be configured with the `mmpmon_path` option in the configuration. If nothing is set, the collector searches in `$PATH` for `mmpmon`. + Metrics: -* `bytes_read` +* `gpfs_bytes_read` * `gpfs_bytes_written` * `gpfs_num_opens` * `gpfs_num_closes` * `gpfs_num_reads` * `gpfs_num_readdirs` * `gpfs_num_inode_updates` +* `gpfs_bw_read` (if `send_bandwidths == true`) +* `gpfs_bw_write` (if `send_bandwidths == true`) The collector adds a `filesystem` tag to all metrics diff --git a/collectors/infinibandMetric.go b/collectors/infinibandMetric.go index ac79e0a..5be095d 100644 --- a/collectors/infinibandMetric.go +++ b/collectors/infinibandMetric.go @@ -16,7 +16,7 @@ import ( "time" ) -const IB_BASEPATH = `/sys/class/infiniband/` +const IB_BASEPATH = "/sys/class/infiniband/" type InfinibandCollectorInfo struct { LID string // IB local Identifier (LID) @@ -24,14 +24,18 @@ type InfinibandCollectorInfo struct { port string // IB device port portCounterFiles map[string]string // mapping counter name -> sysfs file tagSet map[string]string // corresponding tag list + lastState map[string]int64 // State from last measurement } type InfinibandCollector struct { metricCollector config struct { - ExcludeDevices []string `json:"exclude_devices,omitempty"` // IB device to exclude e.g. mlx5_0 + ExcludeDevices []string `json:"exclude_devices,omitempty"` // IB device to exclude e.g. mlx5_0 + SendAbsoluteValues bool `json:"send_abs_values"` // Send absolut values as read from sys filesystem + SendDerivedValues bool `json:"send_derived_values"` // Send derived values e.g. rates } - info []*InfinibandCollectorInfo + info []*InfinibandCollectorInfo + lastTimestamp time.Time // Store time stamp of last tick to derive bandwidths } // Init initializes the Infiniband collector by walking through files below IB_BASEPATH @@ -49,6 +53,11 @@ func (m *InfinibandCollector) Init(config json.RawMessage) error { "source": m.name, "group": "Network", } + + // Set default configuration, + m.config.SendAbsoluteValues = true + m.config.SendDerivedValues = false + // Read configuration file, allow overwriting default config if len(config) > 0 { err = json.Unmarshal(config, &m.config) if err != nil { @@ -60,10 +69,10 @@ func (m *InfinibandCollector) Init(config json.RawMessage) error { globPattern := filepath.Join(IB_BASEPATH, "*", "ports", "*") ibDirs, err := filepath.Glob(globPattern) if err != nil { - return fmt.Errorf("Unable to glob files with pattern %s: %v", globPattern, err) + return fmt.Errorf("unable to glob files with pattern %s: %v", globPattern, err) } if ibDirs == nil { - return fmt.Errorf("Unable to find any directories with pattern %s", globPattern) + return fmt.Errorf("unable to find any directories with pattern %s", globPattern) } for _, path := range ibDirs { @@ -106,10 +115,16 @@ func (m *InfinibandCollector) Init(config json.RawMessage) error { for _, counterFile := range portCounterFiles { err := unix.Access(counterFile, unix.R_OK) if err != nil { - return fmt.Errorf("Unable to access %s: %v", counterFile, err) + return fmt.Errorf("unable to access %s: %v", counterFile, err) } } + // Initialize last state + lastState := make(map[string]int64) + for counter := range portCounterFiles { + lastState[counter] = -1 + } + m.info = append(m.info, &InfinibandCollectorInfo{ LID: LID, @@ -122,11 +137,12 @@ func (m *InfinibandCollector) Init(config json.RawMessage) error { "port": port, "lid": LID, }, + lastState: lastState, }) } if len(m.info) == 0 { - return fmt.Errorf("Found no IB devices") + return fmt.Errorf("found no IB devices") } m.init = true @@ -141,9 +157,17 @@ func (m *InfinibandCollector) Read(interval time.Duration, output chan lp.CCMetr return } + // Current time stamp now := time.Now() + // time difference to last time stamp + timeDiff := now.Sub(m.lastTimestamp).Seconds() + // Save current timestamp + m.lastTimestamp = now + for _, info := range m.info { for counterName, counterFile := range info.portCounterFiles { + + // Read counter file line, err := ioutil.ReadFile(counterFile) if err != nil { cclog.ComponentError( @@ -152,6 +176,8 @@ func (m *InfinibandCollector) Read(interval time.Duration, output chan lp.CCMetr continue } data := strings.TrimSpace(string(line)) + + // convert counter to int64 v, err := strconv.ParseInt(data, 10, 64) if err != nil { cclog.ComponentError( @@ -159,8 +185,24 @@ func (m *InfinibandCollector) Read(interval time.Duration, output chan lp.CCMetr fmt.Sprintf("Read(): Failed to convert Infininiband metrice %s='%s' to int64: %v", counterName, data, err)) continue } - if y, err := lp.New(counterName, info.tagSet, m.meta, map[string]interface{}{"value": v}, now); err == nil { - output <- y + + // Send absolut values + if m.config.SendAbsoluteValues { + if y, err := lp.New(counterName, info.tagSet, m.meta, map[string]interface{}{"value": v}, now); err == nil { + output <- y + } + } + + // Send derived values + if m.config.SendDerivedValues { + if info.lastState[counterName] >= 0 { + rate := float64((v - info.lastState[counterName])) / timeDiff + if y, err := lp.New(counterName+"_bw", info.tagSet, m.meta, map[string]interface{}{"value": rate}, now); err == nil { + output <- y + } + } + // Save current state + info.lastState[counterName] = v } } diff --git a/collectors/infinibandMetric.md b/collectors/infinibandMetric.md index 579ed77..f129aad 100644 --- a/collectors/infinibandMetric.md +++ b/collectors/infinibandMetric.md @@ -5,7 +5,9 @@ "ibstat": { "exclude_devices": [ "mlx4" - ] + ], + "send_abs_values": true, + "send_derived_values": true } ``` @@ -22,5 +24,9 @@ Metrics: * `ib_xmit` * `ib_recv_pkts` * `ib_xmit_pkts` +* `ib_recv_bw` (if `send_derived_values == true`) +* `ib_xmit_bw` (if `send_derived_values == true`) +* `ib_recv_pkts_bw` (if `send_derived_values == true`) +* `ib_xmit_pkts_bw` (if `send_derived_values == true`) The collector adds a `device` tag to all metrics diff --git a/collectors/likwidMetric.go b/collectors/likwidMetric.go index 8ab42d5..e0b0d7e 100644 --- a/collectors/likwidMetric.go +++ b/collectors/likwidMetric.go @@ -15,7 +15,6 @@ import ( "io/ioutil" "math" "os" - "regexp" "strconv" "strings" "time" @@ -28,48 +27,6 @@ import ( "github.com/NVIDIA/go-nvml/pkg/dl" ) -type MetricScope string - -const ( - METRIC_SCOPE_HWTHREAD = iota - METRIC_SCOPE_CORE - METRIC_SCOPE_LLC - METRIC_SCOPE_NUMA - METRIC_SCOPE_DIE - METRIC_SCOPE_SOCKET - METRIC_SCOPE_NODE -) - -func (ms MetricScope) String() string { - return string(ms) -} - -func (ms MetricScope) Likwid() string { - LikwidDomains := map[string]string{ - "cpu": "", - "core": "", - "llc": "C", - "numadomain": "M", - "die": "D", - "socket": "S", - "node": "N", - } - return LikwidDomains[string(ms)] -} - -func (ms MetricScope) Granularity() int { - for i, g := range GetAllMetricScopes() { - if ms == g { - return i - } - } - return -1 -} - -func GetAllMetricScopes() []MetricScope { - return []MetricScope{"cpu" /*, "core", "llc", "numadomain", "die",*/, "socket", "node"} -} - const ( LIKWID_LIB_NAME = "liblikwid.so" LIKWID_LIB_DL_FLAGS = dl.RTLD_LAZY | dl.RTLD_GLOBAL @@ -77,18 +34,16 @@ const ( ) type LikwidCollectorMetricConfig struct { - Name string `json:"name"` // Name of the metric - Calc string `json:"calc"` // Calculation for the metric using - //Aggr string `json:"aggregation"` // if scope unequal to LIKWID metric scope, the values are combined (sum, min, max, mean or avg, median) - Scope MetricScope `json:"scope"` // scope for calculation. subscopes are aggregated using the 'aggregation' function - Publish bool `json:"publish"` - granulatity MetricScope + Name string `json:"name"` // Name of the metric + Calc string `json:"calc"` // Calculation for the metric using + Type string `json:"type"` // Metric type (aka node, socket, cpu, ...) + Publish bool `json:"publish"` + Unit string `json:"unit"` // Unit of metric if any } type LikwidCollectorEventsetConfig struct { - Events map[string]string `json:"events"` - granulatity map[string]MetricScope - Metrics []LikwidCollectorMetricConfig `json:"metrics"` + Events map[string]string `json:"events"` + Metrics []LikwidCollectorMetricConfig `json:"metrics"` } type LikwidCollectorConfig struct { @@ -98,28 +53,28 @@ type LikwidCollectorConfig struct { InvalidToZero bool `json:"invalid_to_zero,omitempty"` AccessMode string `json:"access_mode,omitempty"` DaemonPath string `json:"accessdaemon_path,omitempty"` + LibraryPath string `json:"liblikwid_path,omitempty"` } type LikwidCollector struct { metricCollector - cpulist []C.int - cpu2tid map[int]int - sock2tid map[int]int - scopeRespTids map[MetricScope]map[int]int - metrics map[C.int]map[string]int - groups []C.int - config LikwidCollectorConfig - results map[int]map[int]map[string]interface{} - mresults map[int]map[int]map[string]float64 - gmresults map[int]map[string]float64 - basefreq float64 - running bool + cpulist []C.int + cpu2tid map[int]int + sock2tid map[int]int + metrics map[C.int]map[string]int + groups []C.int + config LikwidCollectorConfig + results map[int]map[int]map[string]interface{} + mresults map[int]map[int]map[string]float64 + gmresults map[int]map[string]float64 + basefreq float64 + running bool } type LikwidMetric struct { name string search string - scope MetricScope + scope string group_idx int } @@ -131,152 +86,43 @@ func eventsToEventStr(events map[string]string) string { return strings.Join(elist, ",") } -func getGranularity(counter, event string) MetricScope { - if strings.HasPrefix(counter, "PMC") || strings.HasPrefix(counter, "FIXC") { - return "cpu" - } else if strings.Contains(counter, "BOX") || strings.Contains(counter, "DEV") { - return "socket" - } else if strings.HasPrefix(counter, "PWR") { - if event == "RAPL_CORE_ENERGY" { - return "cpu" - } else { - return "socket" - } - } - return "unknown" -} - func getBaseFreq() float64 { var freq float64 = math.NaN() C.power_init(0) info := C.get_powerInfo() if float64(info.baseFrequency) != 0 { - freq = float64(info.baseFrequency) * 1e3 + freq = float64(info.baseFrequency) * 1e6 } else { buffer, err := ioutil.ReadFile("/sys/devices/system/cpu/cpu0/cpufreq/bios_limit") if err == nil { data := strings.Replace(string(buffer), "\n", "", -1) x, err := strconv.ParseInt(data, 0, 64) if err == nil { - freq = float64(x) * 1e3 + freq = float64(x) * 1e6 } } } return freq } -func (m *LikwidCollector) initGranularity() { - splitRegex := regexp.MustCompile("[+-/*()]") - for _, evset := range m.config.Eventsets { - evset.granulatity = make(map[string]MetricScope) - for counter, event := range evset.Events { - gran := getGranularity(counter, event) - if gran.Granularity() >= 0 { - evset.granulatity[counter] = gran - } - } - for i, metric := range evset.Metrics { - s := splitRegex.Split(metric.Calc, -1) - gran := MetricScope("cpu") - evset.Metrics[i].granulatity = gran - for _, x := range s { - if _, ok := evset.Events[x]; ok { - if evset.granulatity[x].Granularity() > gran.Granularity() { - gran = evset.granulatity[x] - } - } - } - evset.Metrics[i].granulatity = gran - } - } - for i, metric := range m.config.Metrics { - s := splitRegex.Split(metric.Calc, -1) - gran := MetricScope("cpu") - m.config.Metrics[i].granulatity = gran - for _, x := range s { - for _, evset := range m.config.Eventsets { - for _, m := range evset.Metrics { - if m.Name == x && m.granulatity.Granularity() > gran.Granularity() { - gran = m.granulatity - } - } - } - } - m.config.Metrics[i].granulatity = gran - } -} - -type TopoResolveFunc func(cpuid int) int - -func (m *LikwidCollector) getResponsiblities() map[MetricScope]map[int]int { - get_cpus := func(scope MetricScope) map[int]int { - var slist []int - var cpu C.int - var input func(index int) string - switch scope { - case "node": - slist = []int{0} - input = func(index int) string { return "N:0" } - case "socket": - input = func(index int) string { return fmt.Sprintf("%s%d:0", scope.Likwid(), index) } - slist = topo.SocketList() - // case "numadomain": - // input = func(index int) string { return fmt.Sprintf("%s%d:0", scope.Likwid(), index) } - // slist = topo.NumaNodeList() - // cclog.Debug(scope, " ", input(0), " ", slist) - // case "die": - // input = func(index int) string { return fmt.Sprintf("%s%d:0", scope.Likwid(), index) } - // slist = topo.DieList() - // case "llc": - // input = fmt.Sprintf("%s%d:0", scope.Likwid(), s) - // slist = topo.LLCacheList() - case "cpu": - input = func(index int) string { return fmt.Sprintf("%d", index) } - slist = topo.CpuList() - case "hwthread": - input = func(index int) string { return fmt.Sprintf("%d", index) } - slist = topo.CpuList() - } - outmap := make(map[int]int) - for _, s := range slist { - t := C.CString(input(s)) - clen := C.cpustr_to_cpulist(t, &cpu, 1) - if int(clen) == 1 { - outmap[s] = m.cpu2tid[int(cpu)] - } else { - cclog.Error(fmt.Sprintf("Cannot determine responsible CPU for %s", input(s))) - outmap[s] = -1 - } - C.free(unsafe.Pointer(t)) - } - return outmap - } - - scopes := GetAllMetricScopes() - complete := make(map[MetricScope]map[int]int) - for _, s := range scopes { - complete[s] = get_cpus(s) - } - return complete -} - func (m *LikwidCollector) Init(config json.RawMessage) error { var ret C.int m.name = "LikwidCollector" m.config.AccessMode = LIKWID_DEF_ACCESSMODE + m.config.LibraryPath = LIKWID_LIB_NAME if len(config) > 0 { err := json.Unmarshal(config, &m.config) if err != nil { return err } } - lib := dl.New(LIKWID_LIB_NAME, LIKWID_LIB_DL_FLAGS) + lib := dl.New(m.config.LibraryPath, LIKWID_LIB_DL_FLAGS) if lib == nil { - return fmt.Errorf("error instantiating DynamicLibrary for %s", LIKWID_LIB_NAME) + return fmt.Errorf("error instantiating DynamicLibrary for %s", m.config.LibraryPath) } err := lib.Open() if err != nil { - return fmt.Errorf("error opening %s: %v", LIKWID_LIB_NAME, err) + return fmt.Errorf("error opening %s: %v", m.config.LibraryPath, err) } if m.config.ForceOverwrite { @@ -306,10 +152,6 @@ func (m *LikwidCollector) Init(config json.RawMessage) error { return err } - // Determine which counter works at which level. PMC*: cpu, *BOX*: socket, ... - m.initGranularity() - // Generate map for MetricScope -> scope_id (like socket id) -> responsible id (offset in cpulist) - m.scopeRespTids = m.getResponsiblities() switch m.config.AccessMode { case "direct": C.HPMmode(0) @@ -336,29 +178,36 @@ func (m *LikwidCollector) Init(config json.RawMessage) error { globalParams["inverseClock"] = float64(1.0) // While adding the events, we test the metrics whether they can be computed at all for i, evset := range m.config.Eventsets { - estr := eventsToEventStr(evset.Events) - // Generate parameter list for the metric computing test - params := make(map[string]interface{}) - params["time"] = float64(1.0) - params["inverseClock"] = float64(1.0) - for counter := range evset.Events { - params[counter] = float64(1.0) - } - for _, metric := range evset.Metrics { - // Try to evaluate the metric - _, err := agg.EvalFloat64Condition(metric.Calc, params) - if err != nil { - cclog.ComponentError(m.name, "Calculation for metric", metric.Name, "failed:", err.Error()) - continue + var gid C.int + var cstr *C.char + if len(evset.Events) > 0 { + estr := eventsToEventStr(evset.Events) + // Generate parameter list for the metric computing test + params := make(map[string]interface{}) + params["time"] = float64(1.0) + params["inverseClock"] = float64(1.0) + for counter := range evset.Events { + params[counter] = float64(1.0) } - // If the metric is not in the parameter list for the global metrics, add it - if _, ok := globalParams[metric.Name]; !ok { - globalParams[metric.Name] = float64(1.0) + for _, metric := range evset.Metrics { + // Try to evaluate the metric + _, err := agg.EvalFloat64Condition(metric.Calc, params) + if err != nil { + cclog.ComponentError(m.name, "Calculation for metric", metric.Name, "failed:", err.Error()) + continue + } + // If the metric is not in the parameter list for the global metrics, add it + if _, ok := globalParams[metric.Name]; !ok { + globalParams[metric.Name] = float64(1.0) + } } + // Now we add the list of events to likwid + cstr = C.CString(estr) + gid = C.perfmon_addEventSet(cstr) + } else { + cclog.ComponentError(m.name, "Invalid Likwid eventset config, no events given") + continue } - // Now we add the list of events to likwid - cstr := C.CString(estr) - gid := C.perfmon_addEventSet(cstr) if gid >= 0 { m.groups = append(m.groups, gid) } @@ -434,15 +283,9 @@ func (m *LikwidCollector) calcEventsetMetrics(group int, interval time.Duration, // Go over events and get the results for eidx = 0; int(eidx) < len(evset.Events); eidx++ { ctr := C.perfmon_getCounterName(gid, eidx) - ev := C.perfmon_getEventName(gid, eidx) gctr := C.GoString(ctr) - gev := C.GoString(ev) - // MetricScope for the counter (and if needed the event) - scope := getGranularity(gctr, gev) - // Get the map scope-id -> tids - // This way we read less counters like only the responsible hardware thread for a socket - scopemap := m.scopeRespTids[scope] - for _, tid := range scopemap { + + for _, tid := range m.cpu2tid { if tid >= 0 { m.results[group][tid]["time"] = interval.Seconds() m.results[group][tid]["inverseClock"] = invClock @@ -456,7 +299,10 @@ func (m *LikwidCollector) calcEventsetMetrics(group int, interval time.Duration, for _, metric := range evset.Metrics { // The metric scope is determined in the Init() function // Get the map scope-id -> tids - scopemap := m.scopeRespTids[metric.Scope] + scopemap := m.cpu2tid + if metric.Type == "socket" { + scopemap = m.sock2tid + } for domain, tid := range scopemap { if tid >= 0 { value, err := agg.EvalFloat64Condition(metric.Calc, m.results[group][tid]) @@ -474,13 +320,15 @@ func (m *LikwidCollector) calcEventsetMetrics(group int, interval time.Duration, // Now we have the result, send it with the proper tags if !math.IsNaN(value) { if metric.Publish { - tags := map[string]string{"type": metric.Scope.String()} - if metric.Scope != "node" { - tags["type-id"] = fmt.Sprintf("%d", domain) - } fields := map[string]interface{}{"value": value} - y, err := lp.New(metric.Name, tags, m.meta, fields, time.Now()) + y, err := lp.New(metric.Name, map[string]string{"type": metric.Type}, m.meta, fields, time.Now()) if err == nil { + if metric.Type != "node" { + y.AddTag("type-id", fmt.Sprintf("%d", domain)) + } + if len(metric.Unit) > 0 { + y.AddMeta("unit", metric.Unit) + } output <- y } } @@ -495,7 +343,10 @@ func (m *LikwidCollector) calcEventsetMetrics(group int, interval time.Duration, // Go over the global metrics, derive the value out of the event sets' metric values and send it func (m *LikwidCollector) calcGlobalMetrics(interval time.Duration, output chan lp.CCMetric) error { for _, metric := range m.config.Metrics { - scopemap := m.scopeRespTids[metric.Scope] + scopemap := m.cpu2tid + if metric.Type == "socket" { + scopemap = m.sock2tid + } for domain, tid := range scopemap { if tid >= 0 { // Here we generate parameter list @@ -521,13 +372,16 @@ func (m *LikwidCollector) calcGlobalMetrics(interval time.Duration, output chan // Now we have the result, send it with the proper tags if !math.IsNaN(value) { if metric.Publish { - tags := map[string]string{"type": metric.Scope.String()} - if metric.Scope != "node" { - tags["type-id"] = fmt.Sprintf("%d", domain) - } + tags := map[string]string{"type": metric.Type} fields := map[string]interface{}{"value": value} y, err := lp.New(metric.Name, tags, m.meta, fields, time.Now()) if err == nil { + if metric.Type != "node" { + y.AddTag("type-id", fmt.Sprintf("%d", domain)) + } + if len(metric.Unit) > 0 { + y.AddMeta("unit", metric.Unit) + } output <- y } } diff --git a/collectors/likwidMetric.md b/collectors/likwidMetric.md index 1aa4242..fe28857 100644 --- a/collectors/likwidMetric.md +++ b/collectors/likwidMetric.md @@ -4,14 +4,17 @@ The `likwid` collector is probably the most complicated collector. The LIKWID library is included as static library with *direct* access mode. The *direct* access mode is suitable if the daemon is executed by a root user. The static library does not contain the performance groups, so all information needs to be provided in the configuration. The `likwid` configuration consists of two parts, the "eventsets" and "globalmetrics": -- An event set list itself has two parts, the "events" and a set of derivable "metrics". Each of the "events" is a counter:event pair in LIKWID's syntax. The "metrics" are a list of formulas to derive the metric value from the measurements of the "events". Each metric has a name, the formula, a scope and a publish flag. Counter names can be used like variables in the formulas, so `PMC0+PMC1` sums the measurements for the both events configured in the counters `PMC0` and `PMC1`. The scope tells the Collector whether it is a metric for each hardware thread (`cpu`) or each CPU socket (`socket`). The last one is the publishing flag. It tells the collector whether a metric should be sent to the router. -- The global metrics are metrics which require data from all event set measurements to be derived. The inputs are the metrics in the event sets. Similar to the metrics in the event sets, the global metrics are defined by a name, a formula, a scope and a publish flag. See event set metrics for details. The only difference is that there is no access to the raw event measurements anymore but only to the metrics. So, the idea is to derive a metric in the "eventsets" section and reuse it in the "globalmetrics" part. If you need a metric only for deriving the global metrics, disable forwarding of the event set metrics. **Be aware** that the combination might be misleading because the "behavior" of a metric changes over time and the multiple measurements might count different computing phases. +- An event set list itself has two parts, the "events" and a set of derivable "metrics". Each of the "events" is a counter:event pair in LIKWID's syntax. The "metrics" are a list of formulas to derive the metric value from the measurements of the "events". Each metric has a name, the formula, a scope and a publish flag. Counter names can be used like variables in the formulas, so `PMC0+PMC1` sums the measurements for the both events configured in the counters `PMC0` and `PMC1`. The scope tells the Collector whether it is a metric for each hardware thread (`cpu`) or each CPU socket (`socket`). You may specify a unit for the metric with `unit`. The last one is the publishing flag. It tells the collector whether a metric should be sent to the router. +- The global metrics are metrics which require data from all event set measurements to be derived. The inputs are the metrics in the event sets. Similar to the metrics in the event sets, the global metrics are defined by a name, a formula, a scope and a publish flag. See event set metrics for details. The only difference is that there is no access to the raw event measurements anymore but only to the metrics. So, the idea is to derive a metric in the "eventsets" section and reuse it in the "globalmetrics" part. If you need a metric only for deriving the global metrics, disable forwarding of the event set metrics (`publish=false`). **Be aware** that the combination might be misleading because the "behavior" of a metric changes over time and the multiple measurements might count different computing phases. Similar to the metrics in the eventset, you can specify a metric unit with the `unit` field. Additional options: - `access_mode` : Method to use for hardware performance monitoring (`direct` access as root user, `accessdaemon` for the daemon mode) - `accessdaemon_path`: Folder with the access daemon `likwid-accessD`, commonly `$LIKWID_INSTALL_LOC/sbin` - `force_overwrite`: Same as setting `LIKWID_FORCE=1`. In case counters are already in-use, LIKWID overwrites their configuration to do its measurements - `invalid_to_zero`: In some cases, the calculations result in `NaN` or `Inf`. With this option, all `NaN` and `Inf` values are replaces with `0.0`. +- `access_mode`: Specify LIKWID access mode: `direct` for direct register access as root user or `accessdaemon` +- `accessdaemon_path`: Folder of the accessDaemon `likwid-accessD` +- `liblikwid_path`: Location of `liblikwid.so` ### Available metric scopes @@ -54,7 +57,8 @@ $ scripts/likwid_perfgroup_to_cc_config.py ICX MEM_DP "calc": "time", "name": "Runtime (RDTSC) [s]", "publish": true, - "scope": "hwthread" + "unit": "seconds" + "scope": "cpu" }, { "..." : "..." @@ -104,25 +108,28 @@ $ chwon $CCUSER /var/run/likwid.lock { "name": "ipc", "calc": "PMC0/PMC1", - "scope": "cpu", + "type": "cpu", "publish": true }, { "name": "flops_any", "calc": "0.000001*PMC2/time", - "scope": "cpu", + "unit": "MFlops/s", + "type": "cpu", "publish": true }, { - "name": "clock_mhz", + "name": "clock", "calc": "0.000001*(FIXC1/FIXC2)/inverseClock", - "scope": "cpu", + "type": "cpu", + "unit": "MHz", "publish": true }, { "name": "mem1", "calc": "0.000001*(DFC0+DFC1+DFC2+DFC3)*64.0/time", - "scope": "socket", + "unit": "Mbyte/s", + "type": "socket", "publish": false } ] @@ -140,19 +147,22 @@ $ chwon $CCUSER /var/run/likwid.lock { "name": "pwr_core", "calc": "PWR0/time", - "scope": "socket", + "unit": "Watt" + "type": "socket", "publish": true }, { "name": "pwr_pkg", "calc": "PWR1/time", - "scope": "socket", + "type": "socket", + "unit": "Watt" "publish": true }, { "name": "mem2", "calc": "0.000001*(DFC0+DFC1+DFC2+DFC3)*64.0/time", - "scope": "socket", + "unit": "Mbyte/s", + "type": "socket", "publish": false } ] @@ -162,7 +172,8 @@ $ chwon $CCUSER /var/run/likwid.lock { "name": "mem_bw", "calc": "mem1+mem2", - "scope": "socket", + "type": "socket", + "unit": "Mbyte/s", "publish": true } ] @@ -198,3 +209,4 @@ IPC PMC0/PMC1 -> { -> ] ``` +The script `scripts/likwid_perfgroup_to_cc_config.py` might help you. \ No newline at end of file diff --git a/collectors/lustreMetric.go b/collectors/lustreMetric.go index 66fd3fd..67efd7a 100644 --- a/collectors/lustreMetric.go +++ b/collectors/lustreMetric.go @@ -19,20 +19,23 @@ const LCTL_CMD = `lctl` const LCTL_OPTION = `get_param` type LustreCollectorConfig struct { - LCtlCommand string `json:"lctl_command"` - ExcludeMetrics []string `json:"exclude_metrics"` - SendAllMetrics bool `json:"send_all_metrics"` - Sudo bool `json:"use_sudo"` + LCtlCommand string `json:"lctl_command"` + ExcludeMetrics []string `json:"exclude_metrics"` + SendAllMetrics bool `json:"send_all_metrics"` + Sudo bool `json:"use_sudo"` + SendAbsoluteValues bool `json:"send_abs_values"` + SendDerivedValues bool `json:"send_derived_values"` } type LustreCollector struct { metricCollector - tags map[string]string - matches map[string]map[string]int - stats map[string]map[string]int64 - config LustreCollectorConfig - lctl string - sudoCmd string + tags map[string]string + matches map[string]map[string]int + stats map[string]map[string]int64 + config LustreCollectorConfig + lctl string + sudoCmd string + lastTimestamp time.Time // Store time stamp of last tick to derive bandwidths } func (m *LustreCollector) getDeviceDataCommand(device string) []string { @@ -165,6 +168,7 @@ func (m *LustreCollector) Init(config json.RawMessage) error { } } } + m.lastTimestamp = time.Now() m.init = true return nil } @@ -173,6 +177,8 @@ func (m *LustreCollector) Read(interval time.Duration, output chan lp.CCMetric) if !m.init { return } + now := time.Now() + tdiff := now.Sub(m.lastTimestamp) for device, devData := range m.stats { stats := m.getDeviceDataCommand(device) processed := []string{} @@ -183,23 +189,35 @@ func (m *LustreCollector) Read(interval time.Duration, output chan lp.CCMetric) if fields, ok := m.matches[lf[0]]; ok { for name, idx := range fields { x, err := strconv.ParseInt(lf[idx], 0, 64) - if err != nil { - continue - } - value := x - devData[name] - devData[name] = x - if value < 0 { - value = 0 - } - y, err := lp.New(name, m.tags, m.meta, map[string]interface{}{"value": value}, time.Now()) if err == nil { - y.AddTag("device", device) - if strings.Contains(name, "byte") { - y.AddMeta("unit", "Byte") + value := x - devData[name] + devData[name] = x + if value < 0 { + value = 0 } - output <- y - if m.config.SendAllMetrics { - processed = append(processed, name) + if m.config.SendAbsoluteValues { + y, err := lp.New(name, m.tags, m.meta, map[string]interface{}{"value": value}, time.Now()) + if err == nil { + y.AddTag("device", device) + if strings.Contains(name, "byte") { + y.AddMeta("unit", "Byte") + } + output <- y + if m.config.SendAllMetrics { + processed = append(processed, name) + } + } + } + if m.config.SendDerivedValues && strings.Contains(name, "bytes") { + y, err := lp.New(name+"_bw", m.tags, m.meta, map[string]interface{}{"value": float64(value) / tdiff.Seconds()}, time.Now()) + if err == nil { + y.AddTag("device", device) + y.AddMeta("unit", "Bytes/sec") + output <- y + if m.config.SendAllMetrics { + processed = append(processed, name) + } + } } } } @@ -221,6 +239,7 @@ func (m *LustreCollector) Read(interval time.Duration, output chan lp.CCMetric) } } } + m.lastTimestamp = now } func (m *LustreCollector) Close() { diff --git a/collectors/lustreMetric.md b/collectors/lustreMetric.md index 0cb9fc8..de4ed60 100644 --- a/collectors/lustreMetric.md +++ b/collectors/lustreMetric.md @@ -9,21 +9,26 @@ "exclude_metrics": [ "setattr", "getattr" - ] + ], + "send_abs_values" : true, + "send_derived_values" : true } ``` The `lustrestat` collector reads from the procfs stat files for Lustre like `/proc/fs/lustre/llite/lnec-XXXXXX/stats`. Metrics: -* `read_bytes` -* `read_requests` -* `write_bytes` -* `write_requests` -* `open` -* `close` -* `getattr` -* `setattr` -* `statfs` -* `inode_permission` +* `lustre_read_bytes` +* `lustre_read_requests` +* `lustre_write_bytes` +* `lustre_write_requests` +* `lustre_open` +* `lustre_close` +* `lustre_getattr` +* `lustre_setattr` +* `lustre_statfs` +* `lustre_inode_permission` +* `lustre_read_bytes_bw` (if `send_derived_values == true`) +* `lustre_write_bytes_bw` (if `send_derived_values == true`) +This collector adds an `device` tag. \ No newline at end of file diff --git a/collectors/netstatMetric.go b/collectors/netstatMetric.go index 7eaa3cf..d171d4b 100644 --- a/collectors/netstatMetric.go +++ b/collectors/netstatMetric.go @@ -13,22 +13,27 @@ import ( lp "github.com/ClusterCockpit/cc-metric-collector/internal/ccMetric" ) -const NETSTATFILE = `/proc/net/dev` +const NETSTATFILE = "/proc/net/dev" type NetstatCollectorConfig struct { - IncludeDevices []string `json:"include_devices"` + IncludeDevices []string `json:"include_devices"` + SendAbsoluteValues bool `json:"send_abs_values"` + SendDerivedValues bool `json:"send_derived_values"` } type NetstatCollectorMetric struct { - index int - lastValue float64 + name string + index int + tags map[string]string + meta map[string]string + meta_rates map[string]string + lastValue int64 } type NetstatCollector struct { metricCollector config NetstatCollectorConfig - matches map[string]map[string]NetstatCollectorMetric - devtags map[string]map[string]string + matches map[string][]NetstatCollectorMetric lastTimestamp time.Time } @@ -36,15 +41,33 @@ func (m *NetstatCollector) Init(config json.RawMessage) error { m.name = "NetstatCollector" m.setup() m.lastTimestamp = time.Now() - m.meta = map[string]string{"source": m.name, "group": "Network"} - m.devtags = make(map[string]map[string]string) - nameIndexMap := map[string]int{ - "net_bytes_in": 1, - "net_pkts_in": 2, - "net_bytes_out": 9, - "net_pkts_out": 10, - } - m.matches = make(map[string]map[string]NetstatCollectorMetric) + + const ( + fieldInterface = iota + fieldReceiveBytes + fieldReceivePackets + fieldReceiveErrs + fieldReceiveDrop + fieldReceiveFifo + fieldReceiveFrame + fieldReceiveCompressed + fieldReceiveMulticast + fieldTransmitBytes + fieldTransmitPackets + fieldTransmitErrs + fieldTransmitDrop + fieldTransmitFifo + fieldTransmitColls + fieldTransmitCarrier + fieldTransmitCompressed + ) + + m.matches = make(map[string][]NetstatCollectorMetric) + + // Set default configuration, + m.config.SendAbsoluteValues = true + m.config.SendDerivedValues = false + // Read configuration file, allow overwriting default config if len(config) > 0 { err := json.Unmarshal(config, &m.config) if err != nil { @@ -52,7 +75,9 @@ func (m *NetstatCollector) Init(config json.RawMessage) error { return err } } - file, err := os.Open(string(NETSTATFILE)) + + // Check access to net statistic file + file, err := os.Open(NETSTATFILE) if err != nil { cclog.ComponentError(m.name, err.Error()) return err @@ -62,23 +87,65 @@ func (m *NetstatCollector) Init(config json.RawMessage) error { scanner := bufio.NewScanner(file) for scanner.Scan() { l := scanner.Text() + + // Skip lines with no net device entry if !strings.Contains(l, ":") { continue } + + // Split line into fields f := strings.Fields(l) + + // Get net device entry dev := strings.Trim(f[0], ": ") + + // Check if device is a included device if _, ok := stringArrayContains(m.config.IncludeDevices, dev); ok { - m.matches[dev] = make(map[string]NetstatCollectorMetric) - for name, idx := range nameIndexMap { - m.matches[dev][name] = NetstatCollectorMetric{ - index: idx, - lastValue: 0, - } + tags := map[string]string{"device": dev, "type": "node"} + meta_unit_byte := map[string]string{"source": m.name, "group": "Network", "unit": "bytes"} + meta_unit_byte_per_sec := map[string]string{"source": m.name, "group": "Network", "unit": "bytes/sec"} + meta_unit_pkts := map[string]string{"source": m.name, "group": "Network", "unit": "packets"} + meta_unit_pkts_per_sec := map[string]string{"source": m.name, "group": "Network", "unit": "packets/sec"} + + m.matches[dev] = []NetstatCollectorMetric{ + { + name: "net_bytes_in", + index: fieldReceiveBytes, + lastValue: -1, + tags: tags, + meta: meta_unit_byte, + meta_rates: meta_unit_byte_per_sec, + }, + { + name: "net_pkts_in", + index: fieldReceivePackets, + lastValue: -1, + tags: tags, + meta: meta_unit_pkts, + meta_rates: meta_unit_pkts_per_sec, + }, + { + name: "net_bytes_out", + index: fieldTransmitBytes, + lastValue: -1, + tags: tags, + meta: meta_unit_byte, + meta_rates: meta_unit_byte_per_sec, + }, + { + name: "net_pkts_out", + index: fieldTransmitPackets, + lastValue: -1, + tags: tags, + meta: meta_unit_pkts, + meta_rates: meta_unit_pkts_per_sec, + }, } - m.devtags[dev] = map[string]string{"device": dev, "type": "node"} } + } - if len(m.devtags) == 0 { + + if len(m.matches) == 0 { return errors.New("no devices to collector metrics found") } m.init = true @@ -89,50 +156,62 @@ func (m *NetstatCollector) Read(interval time.Duration, output chan lp.CCMetric) if !m.init { return } + // Current time stamp now := time.Now() + // time difference to last time stamp + timeDiff := now.Sub(m.lastTimestamp).Seconds() + // Save current timestamp + m.lastTimestamp = now + file, err := os.Open(string(NETSTATFILE)) if err != nil { cclog.ComponentError(m.name, err.Error()) return } defer file.Close() - tdiff := now.Sub(m.lastTimestamp) scanner := bufio.NewScanner(file) for scanner.Scan() { l := scanner.Text() + + // Skip lines with no net device entry if !strings.Contains(l, ":") { continue } + + // Split line into fields f := strings.Fields(l) + + // Get net device entry dev := strings.Trim(f[0], ":") + // Check if device is a included device if devmetrics, ok := m.matches[dev]; ok { - for name, data := range devmetrics { - v, err := strconv.ParseFloat(f[data.index], 64) - if err == nil { - vdiff := v - data.lastValue - value := vdiff / tdiff.Seconds() - if data.lastValue == 0 { - value = 0 - } - data.lastValue = v - y, err := lp.New(name, m.devtags[dev], m.meta, map[string]interface{}{"value": value}, now) - if err == nil { - switch { - case strings.Contains(name, "byte"): - y.AddMeta("unit", "bytes/sec") - case strings.Contains(name, "pkt"): - y.AddMeta("unit", "packets/sec") - } + for i := range devmetrics { + metric := &devmetrics[i] + + // Read value + v, err := strconv.ParseInt(f[metric.index], 10, 64) + if err != nil { + continue + } + if m.config.SendAbsoluteValues { + if y, err := lp.New(metric.name, metric.tags, metric.meta, map[string]interface{}{"value": v}, now); err == nil { output <- y } - devmetrics[name] = data + } + if m.config.SendDerivedValues { + if metric.lastValue >= 0 { + rate := float64(v-metric.lastValue) / timeDiff + if y, err := lp.New(metric.name+"_bw", metric.tags, metric.meta_rates, map[string]interface{}{"value": rate}, now); err == nil { + output <- y + } + } + metric.lastValue = v } } } } - m.lastTimestamp = time.Now() } func (m *NetstatCollector) Close() { diff --git a/collectors/netstatMetric.md b/collectors/netstatMetric.md index 90d8600..424cf77 100644 --- a/collectors/netstatMetric.md +++ b/collectors/netstatMetric.md @@ -5,17 +5,23 @@ "netstat": { "include_devices": [ "eth0" - ] + ], + "send_abs_values" : true, + "send_derived_values" : true } ``` The `netstat` collector reads data from `/proc/net/dev` and outputs a handful **node** metrics. With the `include_devices` list you can specify which network devices should be measured. **Note**: Most other collectors use an _exclude_ list instead of an include list. Metrics: -* `net_bytes_in` (`unit=bytes/sec`) -* `net_bytes_out` (`unit=bytes/sec`) -* `net_pkts_in` (`unit=packets/sec`) -* `net_pkts_out` (`unit=packets/sec`) +* `net_bytes_in` (`unit=bytes`) +* `net_bytes_out` (`unit=bytes`) +* `net_pkts_in` (`unit=packets`) +* `net_pkts_out` (`unit=packets`) +* `net_bytes_in_bw` (`unit=bytes/sec` if `send_derived_values == true`) +* `net_bytes_out_bw` (`unit=bytes/sec` if `send_derived_values == true`) +* `net_pkts_in_bw` (`unit=packets/sec` if `send_derived_values == true`) +* `net_pkts_out_bw` (`unit=packets/sec` if `send_derived_values == true`) The device name is added as tag `device`. diff --git a/scripts/cc-metric-collector.deb.control b/scripts/cc-metric-collector.deb.control new file mode 100644 index 0000000..8f752cd --- /dev/null +++ b/scripts/cc-metric-collector.deb.control @@ -0,0 +1,12 @@ +Package: cc-metric-collector +Version: {VERSION} +Installed-Size: {INSTALLED_SIZE} +Architecture: {ARCH} +Maintainer: thomas.gruber@fau.de +Depends: libc6 (>= 2.2.1) +Build-Depends: debhelper-compat (= 13), git, golang-go +Description: Metric collection daemon from the ClusterCockpit suite +Homepage: https://github.com/ClusterCockpit/cc-metric-collector +Source: cc-metric-collector +Rules-Requires-Root: no + diff --git a/sinks/gangliaCommon.go b/sinks/gangliaCommon.go index b2a1b2c..f92550b 100644 --- a/sinks/gangliaCommon.go +++ b/sinks/gangliaCommon.go @@ -148,10 +148,14 @@ type GangliaMetricConfig struct { Unit string Group string Value string + Name string } func GetCommonGangliaConfig(point lp.CCMetric) GangliaMetricConfig { mname := GangliaMetricRename(point.Name()) + if oldname, ok := point.GetMeta("oldname"); ok { + mname = GangliaMetricRename(oldname) + } for _, group := range CommonGangliaMetrics { for _, metric := range group.Metrics { if metric.Name == mname { @@ -187,6 +191,7 @@ func GetCommonGangliaConfig(point lp.CCMetric) GangliaMetricConfig { Tmax: metric.Tmax, Unit: metric.Unit, Value: valueStr, + Name: GangliaMetricRename(mname), } } } @@ -198,10 +203,15 @@ func GetCommonGangliaConfig(point lp.CCMetric) GangliaMetricConfig { Tmax: 0, Unit: "", Value: "", + Name: "", } } func GetGangliaConfig(point lp.CCMetric) GangliaMetricConfig { + mname := GangliaMetricRename(point.Name()) + if oldname, ok := point.GetMeta("oldname"); ok { + mname = GangliaMetricRename(oldname) + } group := "" if g, ok := point.GetMeta("group"); ok { group = g @@ -254,5 +264,6 @@ func GetGangliaConfig(point lp.CCMetric) GangliaMetricConfig { Tmax: DEFAULT_GANGLIA_METRIC_TMAX, Unit: unit, Value: valueStr, + Name: GangliaMetricRename(mname), } } diff --git a/sinks/gangliaSink.go b/sinks/gangliaSink.go index 22096af..5324123 100644 --- a/sinks/gangliaSink.go +++ b/sinks/gangliaSink.go @@ -39,16 +39,13 @@ func (s *GangliaSink) Write(point lp.CCMetric) error { //var tagsstr []string var argstr []string - // Get metric name - metricname := GangliaMetricRename(point.Name()) - // Get metric config (type, value, ... in suitable format) conf := GetCommonGangliaConfig(point) if len(conf.Type) == 0 { conf = GetGangliaConfig(point) } if len(conf.Type) == 0 { - return fmt.Errorf("metric %s has no 'value' field", metricname) + return fmt.Errorf("metric %q (Ganglia name %q) has no 'value' field", point.Name(), conf.Name) } if s.config.AddGangliaGroup { @@ -70,7 +67,7 @@ func (s *GangliaSink) Write(point lp.CCMetric) error { if s.config.AddTypeToName { argstr = append(argstr, fmt.Sprintf("--name=%s", GangliaMetricName(point))) } else { - argstr = append(argstr, fmt.Sprintf("--name=%s", metricname)) + argstr = append(argstr, fmt.Sprintf("--name=%s", conf.Name)) } argstr = append(argstr, fmt.Sprintf("--slope=%s", conf.Slope)) argstr = append(argstr, fmt.Sprintf("--value=%s", conf.Value)) diff --git a/sinks/influxAsyncSink.go b/sinks/influxAsyncSink.go index 7b38873..a2cb64a 100644 --- a/sinks/influxAsyncSink.go +++ b/sinks/influxAsyncSink.go @@ -6,6 +6,7 @@ import ( "encoding/json" "errors" "fmt" + "time" cclog "github.com/ClusterCockpit/cc-metric-collector/internal/ccLogger" lp "github.com/ClusterCockpit/cc-metric-collector/internal/ccMetric" @@ -26,15 +27,21 @@ type InfluxAsyncSinkConfig struct { // Maximum number of points sent to server in single request. Default 5000 BatchSize uint `json:"batch_size,omitempty"` // Interval, in ms, in which is buffer flushed if it has not been already written (by reaching batch size) . Default 1000ms - FlushInterval uint `json:"flush_interval,omitempty"` + FlushInterval uint `json:"flush_interval,omitempty"` + InfluxRetryInterval string `json:"retry_interval"` + InfluxExponentialBase uint `json:"retry_exponential_base"` + InfluxMaxRetries uint `json:"max_retries"` + InfluxMaxRetryTime string `json:"max_retry_time"` } type InfluxAsyncSink struct { sink - client influxdb2.Client - writeApi influxdb2Api.WriteAPI - errors <-chan error - config InfluxAsyncSinkConfig + client influxdb2.Client + writeApi influxdb2Api.WriteAPI + errors <-chan error + config InfluxAsyncSinkConfig + influxRetryInterval uint + influxMaxRetryTime uint } func (s *InfluxAsyncSink) connect() error { @@ -63,6 +70,11 @@ func (s *InfluxAsyncSink) connect() error { InsecureSkipVerify: true, }, ) + clientOptions.SetMaxRetryInterval(s.influxRetryInterval) + clientOptions.SetMaxRetryTime(s.influxMaxRetryTime) + clientOptions.SetExponentialBase(s.config.InfluxExponentialBase) + clientOptions.SetMaxRetries(s.config.InfluxMaxRetries) + s.client = influxdb2.NewClientWithOptions(uri, auth, clientOptions) s.writeApi = s.client.WriteAPI(s.config.Organization, s.config.Database) ok, err := s.client.Ping(context.Background()) @@ -99,6 +111,33 @@ func NewInfluxAsyncSink(name string, config json.RawMessage) (Sink, error) { // Set default for maximum number of points sent to server in single request. s.config.BatchSize = 100 + s.influxRetryInterval = uint(time.Duration(1) * time.Second) + s.config.InfluxRetryInterval = "1s" + s.influxMaxRetryTime = uint(7 * time.Duration(24) * time.Hour) + s.config.InfluxMaxRetryTime = "168h" + s.config.InfluxMaxRetries = 20 + s.config.InfluxExponentialBase = 2 + + // Default retry intervals (in seconds) + // 1 2 + // 2 4 + // 4 8 + // 8 16 + // 16 32 + // 32 64 + // 64 128 + // 128 256 + // 256 512 + // 512 1024 + // 1024 2048 + // 2048 4096 + // 4096 8192 + // 8192 16384 + // 16384 32768 + // 32768 65536 + // 65536 131072 + // 131072 262144 + // 262144 524288 if len(config) > 0 { err := json.Unmarshal(config, &s.config) @@ -114,6 +153,16 @@ func NewInfluxAsyncSink(name string, config json.RawMessage) (Sink, error) { return nil, errors.New("not all configuration variables set required by InfluxAsyncSink") } + toUint := func(duration string, def uint) uint { + t, err := time.ParseDuration(duration) + if err == nil { + return uint(t.Milliseconds()) + } + return def + } + s.influxRetryInterval = toUint(s.config.InfluxRetryInterval, s.influxRetryInterval) + s.influxMaxRetryTime = toUint(s.config.InfluxMaxRetryTime, s.influxMaxRetryTime) + // Connect to InfluxDB server if err := s.connect(); err != nil { return nil, fmt.Errorf("unable to connect: %v", err) diff --git a/sinks/influxAsyncSink.md b/sinks/influxAsyncSink.md index 286c93c..951d67d 100644 --- a/sinks/influxAsyncSink.md +++ b/sinks/influxAsyncSink.md @@ -18,6 +18,10 @@ The `influxasync` sink uses the official [InfluxDB golang client](https://pkg.go "organization": "myorg", "ssl": true, "batch_size": 200, + "retry_interval" : "1s", + "retry_exponential_base" : 2, + "max_retries": 20, + "max_retry_time" : "168h" } } ``` @@ -31,4 +35,10 @@ The `influxasync` sink uses the official [InfluxDB golang client](https://pkg.go - `password`: Password for basic authentification - `organization`: Organization in the InfluxDB - `ssl`: Use SSL connection -- `batch_size`: batch up metrics internally, default 100 \ No newline at end of file +- `batch_size`: batch up metrics internally, default 100 +- `retry_interval`: Base retry interval for failed write requests, default 1s +- `retry_exponential_base`: The retry interval is exponentially increased with this base, default 2 +- `max_retries`: Maximal number of retry attempts +- `max_retry_time`: Maximal time to retry failed writes, default 168h (one week) + +For information about the calculation of the retry interval settings, see [offical influxdb-client-go documentation](https://github.com/influxdata/influxdb-client-go#handling-of-failed-async-writes) \ No newline at end of file diff --git a/sinks/influxSink.go b/sinks/influxSink.go index 11859b2..ed3bb09 100644 --- a/sinks/influxSink.go +++ b/sinks/influxSink.go @@ -6,6 +6,7 @@ import ( "encoding/json" "errors" "fmt" + "time" cclog "github.com/ClusterCockpit/cc-metric-collector/internal/ccLogger" lp "github.com/ClusterCockpit/cc-metric-collector/internal/ccMetric" @@ -15,21 +16,29 @@ import ( type InfluxSinkConfig struct { defaultSinkConfig - Host string `json:"host,omitempty"` - Port string `json:"port,omitempty"` - Database string `json:"database,omitempty"` - User string `json:"user,omitempty"` - Password string `json:"password,omitempty"` - Organization string `json:"organization,omitempty"` - SSL bool `json:"ssl,omitempty"` - RetentionPol string `json:"retention_policy,omitempty"` + Host string `json:"host,omitempty"` + Port string `json:"port,omitempty"` + Database string `json:"database,omitempty"` + User string `json:"user,omitempty"` + Password string `json:"password,omitempty"` + Organization string `json:"organization,omitempty"` + SSL bool `json:"ssl,omitempty"` + RetentionPol string `json:"retention_policy,omitempty"` + InfluxRetryInterval string `json:"retry_interval"` + InfluxExponentialBase uint `json:"retry_exponential_base"` + InfluxMaxRetries uint `json:"max_retries"` + InfluxMaxRetryTime string `json:"max_retry_time"` + //InfluxMaxRetryDelay string `json:"max_retry_delay"` // It is mentioned in the docs but there is no way to set it } type InfluxSink struct { sink - client influxdb2.Client - writeApi influxdb2Api.WriteAPIBlocking - config InfluxSinkConfig + client influxdb2.Client + writeApi influxdb2Api.WriteAPIBlocking + config InfluxSinkConfig + influxRetryInterval uint + influxMaxRetryTime uint + //influxMaxRetryDelay uint } func (s *InfluxSink) connect() error { @@ -52,6 +61,12 @@ func (s *InfluxSink) connect() error { InsecureSkipVerify: true, }, ) + + clientOptions.SetMaxRetryInterval(s.influxRetryInterval) + clientOptions.SetMaxRetryTime(s.influxMaxRetryTime) + clientOptions.SetExponentialBase(s.config.InfluxExponentialBase) + clientOptions.SetMaxRetries(s.config.InfluxMaxRetries) + s.client = influxdb2.NewClientWithOptions(uri, auth, clientOptions) s.writeApi = s.client.WriteAPIBlocking(s.config.Organization, s.config.Database) ok, err := s.client.Ping(context.Background()) @@ -91,6 +106,13 @@ func NewInfluxSink(name string, config json.RawMessage) (Sink, error) { return nil, err } } + s.influxRetryInterval = uint(time.Duration(1) * time.Second) + s.config.InfluxRetryInterval = "1s" + s.influxMaxRetryTime = uint(7 * time.Duration(24) * time.Hour) + s.config.InfluxMaxRetryTime = "168h" + s.config.InfluxMaxRetries = 20 + s.config.InfluxExponentialBase = 2 + if len(s.config.Host) == 0 || len(s.config.Port) == 0 || len(s.config.Database) == 0 || @@ -99,6 +121,16 @@ func NewInfluxSink(name string, config json.RawMessage) (Sink, error) { return nil, errors.New("not all configuration variables set required by InfluxSink") } + toUint := func(duration string, def uint) uint { + t, err := time.ParseDuration(duration) + if err == nil { + return uint(t.Milliseconds()) + } + return def + } + s.influxRetryInterval = toUint(s.config.InfluxRetryInterval, s.influxRetryInterval) + s.influxMaxRetryTime = toUint(s.config.InfluxMaxRetryTime, s.influxMaxRetryTime) + // Connect to InfluxDB server if err := s.connect(); err != nil { return nil, fmt.Errorf("unable to connect: %v", err) diff --git a/sinks/influxSink.md b/sinks/influxSink.md index bd0f576..a099895 100644 --- a/sinks/influxSink.md +++ b/sinks/influxSink.md @@ -17,6 +17,10 @@ The `influxdb` sink uses the official [InfluxDB golang client](https://pkg.go.de "password" : "examplepw", "organization": "myorg", "ssl": true, + "retry_interval" : "1s", + "retry_exponential_base" : 2, + "max_retries": 20, + "max_retry_time" : "168h" } } ``` @@ -29,4 +33,10 @@ The `influxdb` sink uses the official [InfluxDB golang client](https://pkg.go.de - `user`: Username for basic authentification - `password`: Password for basic authentification - `organization`: Organization in the InfluxDB -- `ssl`: Use SSL connection \ No newline at end of file +- `ssl`: Use SSL connection +- `retry_interval`: Base retry interval for failed write requests, default 1s +- `retry_exponential_base`: The retry interval is exponentially increased with this base, default 2 +- `max_retries`: Maximal number of retry attempts +- `max_retry_time`: Maximal time to retry failed writes, default 168h (one week) + +For information about the calculation of the retry interval settings, see [offical influxdb-client-go documentation](https://github.com/influxdata/influxdb-client-go#handling-of-failed-async-writes) \ No newline at end of file diff --git a/sinks/libgangliaSink.go b/sinks/libgangliaSink.go index 1fc7863..3651584 100644 --- a/sinks/libgangliaSink.go +++ b/sinks/libgangliaSink.go @@ -124,24 +124,21 @@ func (s *LibgangliaSink) Write(point lp.CCMetric) error { return s.cstrCache[key] } - // Get metric name - metricname := GangliaMetricRename(point.Name()) - conf := GetCommonGangliaConfig(point) if len(conf.Type) == 0 { conf = GetGangliaConfig(point) } if len(conf.Type) == 0 { - return fmt.Errorf("metric %s has no 'value' field", metricname) + return fmt.Errorf("metric %q (Ganglia name %q) has no 'value' field", point.Name(), conf.Name) } if s.config.AddTypeToName { - metricname = GangliaMetricName(point) + conf.Name = GangliaMetricName(point) } c_value = C.CString(conf.Value) c_type = lookup(conf.Type) - c_name = lookup(metricname) + c_name = lookup(conf.Name) // Add unit unit := ""