On 11.03.23 17:29, Glenn Holmer wrote:
On 3/9/23 23:52, Michael Bien wrote:
Hi Glenn,
the index is downloaded to tmp in compressed form and then imported into
a lucene index which should be in the netbeans cache folder (about 1.9GB).
The cache location can be changed via the regular
netbeans_default_cachedir property (netbeans.conf or CLI flag).
All temp files should be automatically cleaned up but it is possible
that something remains in tmp if NB is closed during a late stage of the
extraction process - but thats why its in tmp :)
Nope. NetBeans crashes, leaving a 1.6G gzip and a 5.5G directory
containing the partially-extracted data, both in /tmp. Doesn't go away
until the OS decides to do something about it.
there is a JVM shutdown hook for the gzip at least, so even if the
regular post-extraction cleanup fails it should be removed by the JVM on
exit as fallback.
The extraction is actually done by apache maven-indexer which NetBeans
uses an EOL version of, since everything moved to JDK 11 by now. This
involves Lucene and other dependencies so the process is not directly
influenced by NB.
We could however, potentially tell maven-indexer to use a sub folder of
tmp and clean that folder on IOException - just to be sure - as second
fallback.
i'll take a look,
-mbien
---------------------------------------------------------------------
To unsubscribe, e-mail: users-unsubscr...@netbeans.apache.org
For additional commands, e-mail: users-h...@netbeans.apache.org
For further information about the NetBeans mailing lists, visit:
https://cwiki.apache.org/confluence/display/NETBEANS/Mailing+lists
