Closed freb closed 10 years ago
Just to be clear, the archives produced at the same size as the archive produced when i gzip the raw myindex.bulk output. However, the MD5 hashse are different.
I Just test bulk.xz and have the same issues. I also can't get bzip2 to work, but I may be missing a library. for bzip2 i get the following:
IllegalArgumentException[no archive session implementation found for path /tmp/test.bulk.bzip2]
I'm using the latest plugin version: elasticsearch-knapsack-1.3.2.0.jar
Thanks, I will investigate the data flushing. For bzip2, you should use the suffix bz2
Thanks for the reply. Good call on the bz2 suffix. Just tried it and get the same "Compressed file ends unexpectedly" message.
I have added a fix, and a test for bulk archiver.
After exporting to an index.bulk.gz file, I am unable to gunzip the data. It works fine for tar.gz, but there seems to be an issue with bulk. I get the following error:
gzip: myindex.bulk.gz: unexpected end of file