Thursday, June 29th, 2006
I’ve managed to fix the caching problem previously mentioned. I think I solved it the night I wrote about it, but I’m only writing about it now. The code has been in BETA since then and has just moved into production.
The log output now looks more like this:
Observer has noticed that processing has […]
Posted in Development | No Comments »
Tuesday, June 27th, 2006
I found out exactly why the cache was not being cleared. Incorrect code.
< my $CachingFile = $FreshPorts::Config::CachingRoot .
'/cache/' . $category_name . '.' . $port_name;
---
> my $CachingFile = $FreshPorts::Config::CachingRoot .
‘/cache/ports/’ . $category_name . ‘/’ . $port_name .
[…]
Posted in Bug fixes | 2 Comments »
Tuesday, June 27th, 2006
I know we have a caching problem. I suspect I know the cause.
The symptoms of the problem are:
a commit has been processed by the system.
the commit can be viewed via the URL /commit.php?message_id=MESSAGEID
the commit can be seen via index.php or date.php
the commit cannot be view on the port page
I think I found the cause of […]
Posted in Bug fixes | No Comments »
Thursday, June 22nd, 2006
Florent Thoumie mentioned today that FreshPorts should not supply package installation instructions for Restricted ports. Why? This extract from /usr/ports/Mk/bsd.port.mk indicates why
# RESTRICTED - Prevent the distribution of distfiles and packages to
# […]
Posted in Bug fixes | No Comments »
Monday, June 12th, 2006
Lars Engels had an idea. Why not have a place where you can download a port skeleton? You know, the files that make up a port (Makefile, pkg-plist, distinfo, etc).
As Lars put it:
It came to my mind when I tried out the Frenzy live cd yesterday. It does not have the iwi-firmware port […]
Posted in Bacula, Help Wanted!, New ideas | 1 Comment »
Monday, June 12th, 2006
I ran some tests today. Without caching, I got 5.36 transactions per second (tps). If commit history is cached, I get 7.25 tps. What actually happens in a transaction isn’t relevent. What is important is that by caching I get 35% more transaction per second. That is, the system performance […]
Posted in Development | No Comments »
Monday, June 12th, 2006
In addition to the port history caching which I mentioned previously, I recently implemented some changes which will make it easier to cache yet more data. Specifically, the port details (e.g short description, long description).
I’ve been playing with benchmarks/siege to see how many transactions per second I can put through my development server. […]
Posted in Development | No Comments »