DaanHoogland 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_r367914999
##########
File path:
server/src/main/java/com/cloud/network/router/VirtualNetworkApplianceManagerImpl.java
##########
@@ -1252,39 +1254,60 @@ protected void runInContext() {
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");
- updateDbHealthChecksFromRouterResponse(router.getId(),
answer.getMonitoringResults());
-
- // Check failing tests and restart if needed
- if (answer.getFailingChecks().size() > 0 &&
StringUtils.isNotBlank(checkFailsToRestartVr)) {
- s_logger.warn("Found failing checks on router " +
router + ". " +
- "Checking failed health checks to see if
router needs reboot");
- for (String failedCheck :
answer.getFailingChecks()) {
-
ActionEventUtils.onActionEvent(User.UID_SYSTEM, Account.ACCOUNT_ID_SYSTEM,
- Domain.ROOT_DOMAIN,
EventTypes.EVENT_ROUTER_HEALTH_CHECKS,
- "Router " + router.getUuid() + " has
failing check " + failedCheck);
- if
(checkFailsToRestartVr.contains(failedCheck)) {
- s_logger.warn("Health Check Alert: Found
failing check " + failedCheck + " in " +
-
RouterHealthChecksFailuresToRestartVrCK + ", attempting restart of router.");
- recreateRouter(router.getId());
- }
- }
- }
- }
+ List<String> failingChecks = getFailingChecks(router,
answer);
+ handleFailingChecks(router, failingChecks);
}
} catch (final Exception ex) {
s_logger.error("Fail to complete the
FetchRouterHealthChecksResultTask! ", ex);
ex.printStackTrace();
}
}
+
+ private List<String> getFailingChecks(DomainRouterVO router,
GetRouterMonitorResultsAnswer answer) {
+ if (answer == null) {
+ s_logger.warn("Unable to fetch monitor results for router " +
router);
+ updateRouterConnectivityHealthCheck(router.getId(), false,
"Communication failed");
+ return Arrays.asList(CONNECTIVITY_TEST);
+ } 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());
+ return Arrays.asList(CONNECTIVITY_TEST);
+ } else {
+ updateRouterConnectivityHealthCheck(router.getId(), true,
"Successfully fetched data");
+ updateDbHealthChecksFromRouterResponse(router.getId(),
answer.getMonitoringResults());
+ return answer.getFailingChecks();
+ }
+ }
+
+ private void handleFailingChecks(DomainRouterVO router, List<String>
failingChecks) {
+ if (failingChecks == null || failingChecks.size() == 0) {
+ return;
+ }
+
+ String alertMessage = "Health checks failed: " +
failingChecks.size() + " failing checks on router " + router.getUuid();
+ _alertMgr.sendAlert(AlertType.ALERT_TYPE_DOMAIN_ROUTER,
router.getDataCenterId(), router.getPodIdToDeployIn(),
+ alertMessage, alertMessage);
+ s_logger.warn(alertMessage + ". Checking failed health checks to
see if router needs reboot");
+
+ String checkFailsToRecreateVr =
RouterHealthChecksFailuresToRecreateVr.valueIn(router.getDataCenterId());
+ StringBuilder failingChecksEvent = new StringBuilder("Router " +
router.getUuid() + " has failing checks:");
+ boolean recreateRouter = false;
+ for (String failedCheck : failingChecks) {
+ failingChecksEvent.append(" ").append(failedCheck);
+ if (StringUtils.isNotBlank(checkFailsToRecreateVr) &&
checkFailsToRecreateVr.contains(failedCheck)) {
+ recreateRouter = true;
+ }
+ }
+
+ ActionEventUtils.onActionEvent(User.UID_SYSTEM,
Account.ACCOUNT_ID_SYSTEM,
+ Domain.ROOT_DOMAIN, EventTypes.EVENT_ROUTER_HEALTH_CHECKS,
failingChecksEvent.toString());
+
+ if (recreateRouter) {
+ s_logger.warn("Health Check Alert: Found failing checks in " +
+ RouterHealthChecksFailuresToRecreateVrCK + ",
attempting recreating router.");
+ recreateRouter(router.getId());
+ }
Review comment:
do the dance
----------------------------------------------------------------
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:
[email protected]
With regards,
Apache Git Services
