I have been trying to get the trunk running on Textdrive.  I consistently hit the limits for the amount of memory allowed per process and get errors allocating memory.  I notice that if I remove the aggregation sidebars that use caching like flickr, then I am able to get SVN r940 running.  If I get rid of all sidebars and just add the flickr one it will spike memory usage and get killed by Textdrive. 

I run Lighttpd which Textdrive just upgraded to 1.4.11 from 1.4.9.  I am not positive of the version of Typo I had running before.  It was the trunk revision as of March 2nd.  The server is FreeBSD with Ruby 1.8.4 .

I am guessing that this has something to do with the caching that the aggregations do.  Is the caching in memory by default or something?  Any advice for me on where to start troubleshooting the issue?  I tried running WEBrick on the server and it immediately runs out of memory as well.

Sorry for the rambling nature of my email.  English is my primary language, but it has been a loooooooong day  :-)

-Steve

Here is some info from the `top` - I have my "old" site running under " http://www.stevelongdo.com" and the r940 version running at "http://blog.stevelongdo.com"  That is why there are so many ruby processes...

PID USERNAME PRI NICE   SIZE    RES STATE  C   TIME   WCPU    CPU COMMAND
94872 rhesus         20   15 59616K 49728K kserel 0   0:20  0.00%  0.00% ruby
73055 rhesus         20   15 59616K 49684K kserel 0   0:16  0.00%  0.00% ruby
73052 rhesus          4   15 48008K 37972K select 0   0:08  0.00%  0.00% ruby
73050 rhesus          4   10 52580K 42304K select 0   0:07  0.00%  0.00% ruby
98565 rhesus          4   15 53844K 43564K select 0   0:04  0.00%  0.00% ruby
39367 rhesus        106   10  3792K  2376K select 0   0:02  0.00%  0.00% sshd
73049 rhesus        106   10  3660K  2304K select 0   0:00  0.00%  0.00% lighttpd

_______________________________________________
Typo-list mailing list
Typo-list@rubyforge.org
http://rubyforge.org/mailman/listinfo/typo-list

Reply via email to