anuragaw commented on a change in pull request #3575: [WIP DO NOT MERGE] Health check feature for virtual router URL: https://github.com/apache/cloudstack/pull/3575#discussion_r358793792
########## File path: server/src/main/java/com/cloud/network/router/VirtualNetworkApplianceManagerImpl.java ########## @@ -1186,6 +1227,431 @@ protected void pushToUpdateQueue(final List<NetworkVO> networks) throws Interrup } } + protected class FetchRouterHealthChecksResultTask extends ManagedContextRunnable { + public FetchRouterHealthChecksResultTask() { + } + + @Override + protected void runInContext() { + try { + final List<DomainRouterVO> routers = _routerDao.listByStateAndManagementServer(VirtualMachine.State.Running, mgmtSrvrId); + s_logger.debug("Found " + routers.size() + " running routers. "); + + for (final DomainRouterVO router : routers) { + GetRouterMonitorResultsAnswer answer = fetchAndUpdateRouterHealthChecks(router, false); + String checkFailsToRestartVr = RouterHealthChecksFailuresToRestartVr.valueIn(router.getDataCenterId()); + if (answer == null) { + s_logger.warn("Unable to fetch monitor results for router " + router); + updateRouterConnectivityHealthCheck(router.getId(), false, "Communication failed"); + } else if (!answer.getResult()) { + s_logger.warn("Failed to fetch monitor results from router " + router + " with details: " + answer.getDetails()); + updateRouterConnectivityHealthCheck(router.getId(), false, "Failed to fetch results with details: " + answer.getDetails()); + } else { + updateRouterConnectivityHealthCheck(router.getId(), true, "Successfully fetched data"); + parseAndMergeDbHealthChecks(router.getId(), answer.getMonitoringResults()); + + // Check failing tests and restart if needed + if (answer.getFailingChecks().size() > 0 && StringUtils.isNotBlank(checkFailsToRestartVr)) { + s_logger.info("Checking failed health checks to see if router needs reboot"); + for (String failedCheck : answer.getFailingChecks()) { + if (checkFailsToRestartVr.contains(failedCheck)) { + s_logger.info("Found failing health check " + failedCheck + " so restarting router."); + rebootRouter(router.getId(), true); + } + } + } + } + } + } catch (final Exception ex) { + s_logger.error("Fail to complete the FetchRouterHealthChecksResultTask! ", ex); + ex.printStackTrace(); + } + } + } + + private Map<String, Map<String, RouterHealthCheckResultVO>> getHealthChecksFromDb(long routerId) { + List<RouterHealthCheckResultVO> healthChecksList = routerHealthCheckResultDao.getHealthCheckResults(routerId); + Map<String, Map<String, RouterHealthCheckResultVO>> healthCheckResults = new HashMap<>(); + if (healthChecksList.isEmpty()) { + return healthCheckResults; + } + + for (RouterHealthCheckResultVO healthCheck : healthChecksList) { + if (!healthCheckResults.containsKey(healthCheck.getCheckType())) { + healthCheckResults.put(healthCheck.getCheckType(), new HashMap<>()); + } + healthCheckResults.get(healthCheck.getCheckType()).put(healthCheck.getCheckName(), healthCheck); + } + + return healthCheckResults; + } + + private RouterHealthCheckResultVO updateRouterConnectivityHealthCheck(final long routerId, boolean connected, String message) { + boolean newEntry = false; + RouterHealthCheckResultVO connectivityVO = routerHealthCheckResultDao.getRouterHealthCheckResult(routerId, "connectivity", "basic"); + if (connectivityVO == null) { + connectivityVO = new RouterHealthCheckResultVO(routerId, "connectivity", "basic"); + newEntry = true; + } + + connectivityVO.setCheckResult(connected); + connectivityVO.setLastUpdateTime(new Date()); + connectivityVO.setCheckDetails(StringUtils.isNotEmpty(message) ? message.getBytes(Charset.forName("US-ASCII")) : null); + + if (newEntry) { + routerHealthCheckResultDao.persist(connectivityVO); + } else { + routerHealthCheckResultDao.update(connectivityVO.getId(), connectivityVO); + } + + return routerHealthCheckResultDao.getRouterHealthCheckResult(routerId, "connectivity", "basic"); + } + + private List<RouterHealthCheckResult> parseAndMergeDbHealthChecks(final long routerId, final String monitoringResult) { + if (StringUtils.isBlank(monitoringResult)) { + s_logger.warn("Attempted parsing empty monitoring results string for router " + routerId); + return Collections.emptyList(); + } + + try { + s_logger.info("Parsing and updating DB health check data for router: " + routerId); + s_logger.info("Retrieved data" + monitoringResult); + final Map<String, Map<String, RouterHealthCheckResultVO>> checksInDb = getHealthChecksFromDb(routerId); + final Type t = new TypeToken<Map<String, Map<String, Map<String, String>>>>() {}.getType(); + final Map<String, Map<String, Map<String, String>>> checks = GsonHelper.getGson().fromJson(monitoringResult, t); + final String lastRunKey = "lastRun"; + List<RouterHealthCheckResult> healthChecks = new ArrayList<>(); + for (String checkType : checks.keySet()) { + if (checks.get(checkType).containsKey(lastRunKey)) { + Map<String, String> lastRun = checks.get(checkType).get(lastRunKey); + s_logger.info("Found check types executed on VR " + checkType + ", start: " + lastRun.get("start") + + ", end: " + lastRun.get("end") + ", duration: " + lastRun.get("duration")); + } + + for (String checkName : checks.get(checkType).keySet()) { + if (lastRunKey.equals(checkName)) { + continue; + } + Map<String, String> checkData = checks.get(checkType).get(checkName); + try { + boolean success = Boolean.parseBoolean(checkData.get("success")); + Date lastUpdate = new Date(Long.parseLong(checkData.get("lastUpdate"))); + double lastRunDuration = Double.parseDouble(checkData.get("lastRunDuration")); + String message = checkData.get("message"); + final RouterHealthCheckResultVO hcVo; + boolean newEntry = false; + if (checksInDb.containsKey(checkType) && checksInDb.get(checkType).containsKey(checkName)) { + hcVo = checksInDb.get(checkType).get(checkName); + } else { + hcVo = new RouterHealthCheckResultVO(routerId, checkName, checkType); + newEntry = true; + } + + hcVo.setCheckResult(success); + hcVo.setLastUpdateTime(lastUpdate); + hcVo.setCheckDetails(StringUtils.isNotEmpty(message) ? message.getBytes(Charset.forName("US-ASCII")) : null); + if (newEntry) { + routerHealthCheckResultDao.persist(hcVo); + } else { + routerHealthCheckResultDao.update(hcVo.getId(), hcVo); + } + healthChecks.add(hcVo); + s_logger.info("Found health check " + hcVo + " which took running duration (secs) " + lastRunDuration); + } catch (Exception ex) { + s_logger.error("Skipping health check: Exception while parsing check result data for router id " + routerId + + ", check type: " + checkType + ", check name: " + checkName + ":" + ex.getLocalizedMessage()); + ex.printStackTrace(); + } + } + } + return healthChecks; + } catch (JsonSyntaxException ex) { + s_logger.error("Unable to parse the result of health checks due to " + ex.getLocalizedMessage()); + ex.printStackTrace(); + } + + return Collections.emptyList(); + } + + // Returns null if health checks are not enabled + private GetRouterMonitorResultsAnswer fetchAndUpdateRouterHealthChecks(DomainRouterVO router, boolean performFreshChecks) { + if (!RouterHealthChecksEnabled.valueIn(router.getDataCenterId())) { + return null; + } + + String controlIP = getRouterControlIP(router); + if (StringUtils.isNotBlank(controlIP) && !controlIP.equals("0.0.0.0")) { + final GetRouterMonitorResultsCommand command = new GetRouterMonitorResultsCommand(performFreshChecks); + command.setAccessDetail(NetworkElementCommand.ROUTER_IP, controlIP); + command.setAccessDetail(NetworkElementCommand.ROUTER_NAME, router.getInstanceName()); + try { + final Answer answer = _agentMgr.easySend(router.getHostId(), command); + + if (answer == null) { + s_logger.warn("Unable to fetch monitoring results data from router " + router.getHostName()); + return null; + } + if (answer instanceof GetRouterMonitorResultsAnswer) { + return (GetRouterMonitorResultsAnswer) answer; + } else { + s_logger.warn("Unable to fetch health checks results to router " + router.getHostName() + " Received answer " + answer.getDetails()); + return new GetRouterMonitorResultsAnswer(command, false, null, answer.getDetails()); + } + } catch (final Exception e) { + s_logger.warn("Error while collecting alerts from router: " + router.getInstanceName(), e); + return null; + } + } + + return null; + } + + @Override + public boolean performRouterHealthChecks(long routerId) { + DomainRouterVO router = _routerDao.findById(routerId); + + if (router == null) { + throw new CloudRuntimeException("Unable to find router with id " + routerId); + } + + if (!RouterHealthChecksEnabled.valueIn(router.getDataCenterId())) { + throw new CloudRuntimeException("Router health checks are not enabled in cluster router: " + router); + } + + s_logger.info("Running health check results for router " + router.getUuid()); + + // Step 1: Update health check data on router. + if (!updateRouterHealthCheckData(router)) { + s_logger.warn("Unable to update health check data for fresh run successfully for router: " + router); + return false; + } + s_logger.info("Updated health check data for fresh run successfully for router: " + router); + + // Step 2: Perform and retrieve health checks on router + s_logger.info("Retrieving results for fresh health check execution for router " + router.getUuid()); + GetRouterMonitorResultsAnswer answer = fetchAndUpdateRouterHealthChecks(router, true); + + // Step 3: Update health checks values in database + if (answer == null || !answer.getResult()) { + updateRouterConnectivityHealthCheck(routerId, false, + answer == null ? "Communication failed " : "Failed to fetch results with details: " + answer.getDetails()); + } else { + updateRouterConnectivityHealthCheck(routerId, true, "Successfully fetched data"); + parseAndMergeDbHealthChecks(routerId, answer.getMonitoringResults()); + } + + return true; + } + + protected class UpdateRouterHealthChecksConfigDataTask extends ManagedContextRunnable { + public UpdateRouterHealthChecksConfigDataTask() { + } + + @Override + protected void runInContext() { + try { + final List<DomainRouterVO> routers = _routerDao.listByStateAndManagementServer(VirtualMachine.State.Running, mgmtSrvrId); + s_logger.debug("Found " + routers.size() + " running routers. "); + + for (final DomainRouterVO router : routers) { + updateRouterHealthCheckData(router); + } + } catch (final Exception ex) { + s_logger.error("Fail to complete the UpdateRouterHealthChecksConfigDataTask! ", ex); + } + } + } + + private SetMonitorServiceCommand createMonitorServiceCommand(DomainRouterVO router, List<MonitorServiceTO> services, + boolean reconfigure, boolean deleteFromProcessedCache) { + final SetMonitorServiceCommand command = new SetMonitorServiceCommand(services); + command.setAccessDetail(NetworkElementCommand.ROUTER_IP, getRouterControlIP(router)); + command.setAccessDetail(NetworkElementCommand.ROUTER_NAME, router.getInstanceName()); + command.setAccessDetail(SetMonitorServiceCommand.ROUTER_HEALTH_CHECKS_ENABLED, RouterHealthChecksEnabled.valueIn(router.getDataCenterId()).toString()); + command.setAccessDetail(SetMonitorServiceCommand.ROUTER_HEALTH_CHECKS_BASIC_INTERVAL, RouterHealthChecksBasicInterval.value().toString()); + command.setAccessDetail(SetMonitorServiceCommand.ROUTER_HEALTH_CHECKS_ADVANCED_INTERVAL, RouterHealthChecksAdvancedInterval.value().toString()); + command.setAccessDetail(SetMonitorServiceCommand.ROUTER_HEALTH_CHECKS_EXCLUDED, RouterHealthChecksToExclude.valueIn(router.getDataCenterId())); + command.setAdditionalData(getAdditionalDataForRouterHealthChecks(router)); + command.setReconfigureAfterUpdate(reconfigure); + command.setDeleteFromProcessedCache(deleteFromProcessedCache); // As part of updating + return command; + } + + private boolean updateRouterHealthCheckData(DomainRouterVO router) { + if (!RouterHealthChecksEnabled.valueIn(router.getDataCenterId())) { + return false; + } + + SetMonitorServiceCommand command = createMonitorServiceCommand(router, null,true, true); + String controlIP = getRouterControlIP(router); + if (StringUtils.isBlank(controlIP) || controlIP.equals("0.0.0.0")) { + s_logger.warn("Skipping update data on router " + router.getUuid() + " because controlIp is not correct."); + return false; + } + + s_logger.info("Updating data for router health checks for router " + router.getUuid()); + Answer origAnswer = null; + try { + origAnswer = _agentMgr.easySend(router.getHostId(), command); + } catch (final Exception e) { + s_logger.error("Error while sending update data for health check to router: " + router.getInstanceName(), e); + return false; + } + + if (origAnswer == null) { + s_logger.error("Unable to update health checks data to router " + router.getHostName()); + return false; + } + + GroupAnswer answer = null; + if (origAnswer instanceof GroupAnswer) { + answer = (GroupAnswer) origAnswer; + } else { + s_logger.error("Unable to update health checks data to router " + router.getHostName() + " Received answer " + origAnswer.getDetails()); + return false; + } + + if (!answer.getResult()) { + s_logger.warn("Unable to update health checks data to router " + router.getHostName() + ", details : " + answer.getDetails()); + } + + return answer.getResult(); Review comment: This command pushes data to router and returns true/false. Any error/exception is logged already. ---------------------------------------------------------------- This is an automated message from the Apache Git Service. To respond to the message, please log on to GitHub and use the URL above to go to the specific comment. For queries about this service, please contact Infrastructure at: us...@infra.apache.org With regards, Apache Git Services