There is some experience on backup to s3
https://issues.apache.org/jira/browse/SOLR-9952 iirc, it lacks
performance.
Jörn, it's not a point, but literally s3 consistency might be enough, since
s3 provides read-after-write for PUT and Lucene index writer is
append-only.
On Thu, Jan 17, 2019 at 10:1
This is not a requirement. This is a statement to a problem where there could
be other solutions. s3 is only eventually consistent and I am not sure Solr
works properly in this case. You may also need to check the S3 consistency to
be applied.
> Am 16.01.2019 um 19:39 schrieb Naveen M :
>
> hi
Theoretically you should be able to use the HDFS backend, which you can
configure to use s3. Last time I tried that it did however not work for
some reason. Here is an example for that, which also seems to have
ultimately failed:
https://community.plm.automation.siemens.com/t5/Developer-Space/R
