in webindex/modules/data/src/main/java/webindex/data/LoadS3.java [47:104]
public static void main(String[] args) throws Exception {
if (args.length != 2) {
log.error("Usage: LoadS3 <pathsFile> <range>");
System.exit(1);
}
final List<String> loadList = IndexEnv.getPathsRange(args[0], args[1]);
if (loadList.isEmpty()) {
log.error("No files to load given {} {}", args[0], args[1]);
System.exit(1);
}
final WebIndexConfig webIndexConfig = WebIndexConfig.load();
final int rateLimit = webIndexConfig.getLoadRateLimit();
final String appName = webIndexConfig.fluoApp;
SparkConf sparkConf = new SparkConf().setAppName("webindex-load-s3");
try (JavaSparkContext ctx = new JavaSparkContext(sparkConf)) {
log.info("Loading {} files (Range {} of paths file {}) from AWS", loadList.size(), args[1],
args[0]);
JavaRDD<String> loadRDD = ctx.parallelize(loadList, loadList.size());
final String prefix = WebIndexConfig.CC_URL_PREFIX;
loadRDD.foreachPartition(iter -> {
final FluoConfiguration fluoConfig =
new FluoConfiguration(new File("fluo-conn.properties"));
fluoConfig.setApplicationName(appName);
final RateLimiter rateLimiter = rateLimit > 0 ? RateLimiter.create(rateLimit) : null;
try (FluoClient client = FluoFactory.newClient(fluoConfig);
LoaderExecutor le = client.newLoaderExecutor()) {
iter.forEachRemaining(path -> {
String urlToCopy = prefix + path;
log.info("Loading {} to Fluo", urlToCopy);
try {
ArchiveReader reader = WARCReaderFactory.get(new URL(urlToCopy), 0);
for (ArchiveRecord record : reader) {
Page page = ArchiveUtil.buildPageIgnoreErrors(record);
if (page.getOutboundLinks().size() > 0) {
log.info("Loading page {} with {} links", page.getUrl(),
page.getOutboundLinks().size());
if (rateLimiter != null) {
rateLimiter.acquire();
}
le.execute(PageLoader.updatePage(page));
}
}
} catch (Exception e) {
log.error("Exception while processing {}", path, e);
}
});
}
});
}
}