While providing a custom logback.xml do remember to add Sling specific handlers [1]. Otherwise OSGi integration of logging config would not work as expected
Chetan Mehrotra [1] https://sling.apache.org/documentation/development/logging.html#external-config-file On Thu, Sep 14, 2017 at 6:45 AM, John Logan <john.lo...@texture.com> wrote: > I got this working and thought I'd follow up with what I did in case anyone > else needs this sort of thing. > > > I used SiftingAppender pretty much as shown in any of the examples one can > find online. I put the following logback.xml in my sling.home, and pointed > the OSGI configuration variable > org.apache.sling.commons.log.configurationFile to it: > > > <configuration> > <appender name="SIFT" class="ch.qos.logback.classic.sift.SiftingAppender"> > <discriminator> > <key>jobId</key> > <defaultValue>/var/log/sling/error.log</defaultValue> > </discriminator> > <sift> > <appender name="FILE-${jobId}" class="ch.qos.logback.core.FileAppender"> > <file>${logPath}</file> > <layout class="ch.qos.logback.classic.PatternLayout"> > <pattern>%d{HH:mm:ss.SSS} [%thread] %-5level %logger{36} - > %msg%n</pattern> > </layout> > </appender> > </sift> > </appender> > > <root level="info"> > <appender-ref ref="SIFT" /> > </root> > </configuration> > > > An abstract wrapper implementation of JobExecutor sets up the MDC to conform > to what's expected in the logback.xml, and provides a few other niceties. > Replace the SlingResourceProvider/StorageNode stuff with whatever you use to > access files and Sling nodes; it's just a simple abstraction layer that we > happen to be using. > > > package com.xyz.content.sling.processor; > > import java.io.BufferedReader; > import java.io.IOException; > import java.nio.charset.StandardCharsets; > import java.nio.file.Files; > import java.nio.file.Path; > import java.nio.file.Paths; > > import org.apache.sling.api.resource.LoginException; > import org.apache.sling.api.resource.ResourceResolver; > import org.apache.sling.api.resource.ResourceResolverFactory; > import org.apache.sling.event.jobs.Job; > import org.apache.sling.event.jobs.consumer.JobExecutionContext; > import org.apache.sling.event.jobs.consumer.JobExecutionResult; > import org.apache.sling.event.jobs.consumer.JobExecutor; > import org.slf4j.Logger; > import org.slf4j.LoggerFactory; > import org.slf4j.MDC; > > import com.xyz.content.sling.storage.SlingResourceProvider; > import com.xyz.storage.StorageNode; > > /** > * Sling job executor with build job logging. > * > * This class wraps build processing with job-specific log management. > * Job submitters need to configure the following job properties > * prior to submitting a job: > * * jobId - A symbolic identifier for the job. This value must be > * unique throughout the life of the job. Subclasses > * may use this value as a job name for submitting > * and monitoring SLURM jobs. > * * logPath - The pathname to the build log file. > * * damLogPath - If specified, the DAM resource to which > * the log output should be copied upon build completion. > * * clearLog - An option boolean parameter which, when true, > * removes the build log file if it exists prior > * to commencing the build. > * > * @author john > * > */ > public abstract class BuildJobExecutor implements JobExecutor { > private static Logger LOG = > LoggerFactory.getLogger(BuildJobExecutor.class); > > /** > * Retrieve a resource resolver factory for build processing. > * > * @return the resource resolver factory > */ > protected abstract ResourceResolverFactory getResolverFactory(); > > /** > * Subclass-specific build processing method. > * > * @param job > * @param context > * @param resolver > * @return the result of build processing > */ > protected abstract JobExecutionResult build(Job job, JobExecutionContext > context, ResourceResolver resolver); > > @Override > public JobExecutionResult process(Job job, JobExecutionContext context) { > // > // Prepare for the log directory and file. > // > final String jobId = job.getProperty("jobId", String.class); > final Path logPath = Paths.get(job.getProperty("logPath", > String.class)); > final Path logParentPath = logPath.getParent(); > if (logParentPath != null) { > try { > Files.createDirectories(logParentPath); > } > catch (final IOException e) { > return handleError(context, "Unable to create log directory " > + logParentPath); > } > } > > if (Boolean.TRUE.equals(job.getProperty("resetLog"))) { > try { > Files.deleteIfExists(logPath); > } > catch (final IOException e) { > return handleError(context, "Unable to clear log file " + > logPath); > } > } > > LOG.info("Starting build job with ID " + jobId); > ResourceResolver resolver; > try { > resolver = getResolverFactory().getServiceResourceResolver(null); > } > catch (final LoginException e) { > return handleError(context, "Unable get build job resource > resolver, check cbservice user configuration."); > } > > // > // Perform the build operation. Logging to the job-specific log > file starts here. > // > MDC.put("jobId", jobId); > MDC.put("logPath", logPath.toString()); > try { > final JobExecutionResult result = build(job, context, resolver); > if (job.getProperty("damLogPath") != null) { > final Path damLogPath = > Paths.get(job.getProperty("damLogPath", String.class)); > saveLog(resolver, logPath, damLogPath); > } > return result; > } > catch (final Throwable e) { > LOG.error("Build job failed with an exception.", e); > return handleError(context, "Build job failed with an exception: > " + e.getMessage()); > } > finally { > MDC.remove("jobId"); > MDC.remove("logPath"); > resolver.close(); > } > } > > private JobExecutionResult handleError(JobExecutionContext context, > String message) { > LOG.error(message); > return context.result().message(message).cancelled(); > } > > private void saveLog(ResourceResolver resolver, Path logPath, Path > damLogPath) { > try (BufferedReader reader = Files.newBufferedReader(logPath, > StandardCharsets.UTF_8)) { > final SlingResourceProvider storageProvider = new > SlingResourceProvider(); > storageProvider.setResolver(resolver); > final StorageNode damLogNode = storageProvider.get(damLogPath); > damLogNode.copyFromReader(reader, StandardCharsets.UTF_8); > resolver.commit(); > } > catch (final IOException e) { > LOG.error("Unable to move build log " + logPath + " to DAM > resource " + damLogPath, e); > } > } > } > > > > ________________________________ > From: Robert Munteanu <romb...@apache.org> > Sent: Friday, September 8, 2017 1:13:02 AM > To: users@sling.apache.org > Subject: Re: Directing Sling job logging output to separate files? > > Hi John, > > On Fri, 2017-09-08 at 05:28 +0000, John Logan wrote: >> Hi, >> >> >> I'm using the Sling job manager to handle some long running tasks, >> and would like to direct the log output for each job to its own file >> at a job-specific path. Is there a straightforward way to achieve >> this? > > If your jobs use separate loggers, you can achieve that either by: > > - manually creating loggers and appenders via http://localhost:8080/sys > tem/console/slinglog/ > - adding specific loggers/appenders to the provisioning model > > There might be a way of adding those at runtime using the logback APIs, > but I haven't tried it before. > > Robert
