Re: [Archive-access-discuss] GzippedInputStream error

[Erik H. <eri...@uc...>]

At Fri, 20 Jan 2012 17:11:45 +0100,
raffaele messuti wrote:
> 
> i got this error trying to enqueue some warcs into wayback (1.6)
> 
> ➜   ./bin/cdx-indexer data/warcs/jlis-20012010.warc.gz 
> java.io.IOException: Resetting to invalid mark
> 	at java.io.BufferedInputStream.reset(BufferedInputStream.java:416)
> 	at org.archive.io.GzippedInputStream.<init>(GzippedInputStream.java:123)
> 	at org.archive.io.GzippedInputStream.<init>(GzippedInputStream.java:84)
> 	at org.archive.io.warc.WARCReaderFactory$CompressedWARCReader.<init>(WARCReaderFactory.java:221)
> 	at org.archive.io.warc.WARCReaderFactory.getArchiveReader(WARCReaderFactory.java:88)
> 	at org.archive.io.ArchiveReaderFactory.getArchiveReader(ArchiveReaderFactory.java:110)
> 	at org.archive.io.warc.WARCReaderFactory.get(WARCReaderFactory.java:63)
> 	at org.archive.wayback.resourcestore.indexer.WarcIndexer.iterator(WarcIndexer.java:71)
> 	at org.archive.wayback.resourcestore.indexer.IndexWorker.indexFile(IndexWorker.java:135)
> 	at org.archive.wayback.resourcestore.indexer.IndexWorker.main(IndexWorker.java:204)
> 
> 
> warcs are made with wget-warc,
> my current java version is "1.6.0_29"
> 
> i tested with another java version (1.6.0_21) and worked
> 
> is guess is something related with https://webarchive.jira.com/browse/HER-1865

Hi Raffaele,

Yes, that is almost certainly the issue. The solution is to use the
old JDK version (that is what we are doing at CDL) or upgrade wayback
to 1.6.1 (not yet released). See this message from Brad Tofel:

  http://sourceforge.net/mailarchive/forum.php?thread_name=CCCA2F48128C1F4DAC8B38F9B49C50BE071D53%40OLADAGQP.lao.ola.org&forum_name=archive-access-discuss

best, Erik