Harvest

Good morning,

In my organization, we have both RA and Harvest. When we have an error in RA, there are many types of logs available where I can check errors while executing actions, connectivity, etc.

With harvest, i know the logs are in /opt/CA/scm/logs, but I only see the XXXXHBroker.log and XXXXHServer.log. Do those files save info about errors, such as sudden application fails and such? We recently had an issue where harvest closed while doing a check in (it ended up being due to an out or memory error), but we could not find info about the crashes in those.

Which client have you used to perform check-in operation Workbench or Command Line Utility?

Harvest will log SCM transaction information from the client to the broker within the HBroker log and then the HServer

receiving the transaction request from the Broker will log information from the client. Since both the Broker and Servers are executing on the Harvest server then most likely application failures at the client level will never reach the server and therefore will not have any information to record. The Harvest broker manages the pool of hservers, and it is the hserver that actually processes the client transaction request with the database and returns the results to the client.

The harvest command line utilities produce local log files that will most likely log local failure information.

Workbench will also create a local log file in your workspace on the client machine

(eg: (c:\Users\<yourprofile>\.cascm\workspace\.metadata\cascm.log).

Also we have found the windows system and application event logs have proven useful in providing local application error information. If the Harvest application failed to log any application failure then that is because the application failure occurred before any logging code was executed, so in those cases we look into the windows logs.

In regards to an "out of memory" error, if it occurred within the local machine then the user should be aware of current running processes taking up local memory, try and reduce this workload as much as possible in order to free up more memory. Sometimes increasing hardware memory is the solution.