Hi there,

There is a way to query the short log from the api using:

Slaves.GetSlaveReports()

and a way to get the contents of ALL slave reports:

Slaves.GetSlaveReportsContents()

But i could not find a way to simply figure out where the log files live for a particular slave. Could you guys point me in the right direction?
In particular, looking for these entries - organized by slaves:

\inferno2\deadline\repository8\reports\slaves\01\1\56b171a23284a335683f2011.bz2

(we are aggregating reports for statistics purposes, and would need to be able to do finer control than simply getting all logs at once)

cheers
laszlo

A far as I know you won’t find them in repository. The slave logs are stored on each node separately.

source: http://docs.thinkboxsoftware.com/products/deadline/9.0/1_User%20Manual/manual/logs.html#application-logs

As long as the job is still on the farm, then we found going through the rest api, gave us the quickest way to get at this data:

	#parse url() does the following:
	#response = urllib2.urlopen(url)
        data = json.load(response)
	# 
	idx = 0
	# so you will want something like this if you have events that runs on a job or task event, as they will show up here
        if not include_events:
            # now we need to find the first index, that is not an event log
            get_url = "http://{0}:8080/api/taskreports?Data=all&JobID={1}&TaskID={2}".format(webservicesnode,
                                                                                                        jobid, taskid)
            result = self.parse_url(get_url)
            
            if result:
                idx = 0
                # go throught tasks from newest to last
                for log_d in result:
                    if self.debug: print 'result line::',log_d
                    if not 'Event Log' in log_d.get('Title'):
                        #only break out if we are no longer a event line
                        break
                    idx = idx + 1

                log_result = idx
        #
        get_url = "http://{0}:8080/api/taskreports?Data=alllogcontents&JobID={1}&TaskID={2}".format(webservicesnode, jobid, taskid)
        result = self.parse_url(get_url)
        # this is a list
        #print 'result:',pprint(result)
        # the first entry is the most recent
        if not len(result) == 0:

            last_log = result[log_result]
            # do stuff with the log here

If anyone has a faster way to do this im all ears.

Cheers
Kym

Boop:

job_id = " <job id goodness> "
reports = RepositoryUtils.GetJobReports(job_id)
for report in reports:
    report = RepositoryUtils.GetJobReportLogFileName(report)
    print("Your report is here: {0}".format(report))

Written by hand, may not quite parse properly.

These are for the Job logs, I was asking about the Slave logs. The ones you find here:

repository\reports\slaves\

Ya drive a hard bargain amigo. Here you go:

slave_name = " <some slave name> "
reports = RepositoryUtils.GetSlaveReports(slave_name).GetSlaveReports() // Yeah, the repetition is a thing
for report in reports:
    report = RepositoryUtils.GetSlaveReportLogFileName(report)
    print("Your report is here: {0}".format(report))