From: Neil Bingham <neil@eh...> - 2004-02-16 15:28:57
I've been a long time user of LXR 0.3. Recently we decided to give 0.9.x a
crack since there were some nice new features we wanted to take advantage of.
Our CVS repository is rather large - occupying some 40GB of space and 8 years
of development work.
The problem we are having is that genxref dies with an out of memory error
after running for about a day.
The hardware isn't too shabby;
Compaq Proliant ML370 1GHz with 1GB of RAM. After these errors I upped the
amount of swap from 2 to 10GB and tried again. Same result. Last time I saw
the output from top the process was eating 4GB of memory.
Can anyone suggest any reason why this would be taking this amount of RAM, or
another way to generate the indexes so that it doesn't fall over?
I thought of indexing a top level directory at a time of the repository but
then it wouldn't cross reference between the directories would it?