We are also storing uploaded content to S3. But we are using python and the Botox library for that. We are using asynchronous procedure like you. I don't know joomla technically but I know there is a wordpress module (PHP like Joomla) that does exactly what you are expecting but for wordpress. You should have a look at it. When file are stored on S3 app roles can scale without any problem.
Envoyé de mon iPhone Le 25 juil. 2009 à 00:26, jdMelton <[email protected]> a écrit : > > Jeremy, > > I manage a Scalr.net account where there are multiple websites and one > of them uses Joomla. The farm uses www, app, and mysqllvm roles. The > load on the farm fluctuates so a new app server gets created a few > times per week. So I decided to store files in S3. However, I did not > want the S3 files to be open to the public. I wanted them served from > S3 through the app role to avoid additional bandwidth fees and to > prevent unauthorized access since some of the files are marked private > by the user. > > Since Joomla stores uploaded files (mainly images, text, and pdf > files) in a known parent directory tree, it seemed the easiest thing > to do was copy the directories periodically to S3. Then use a > rewritemap to check for the files in S3 if they were not found > locally. > > I was concerned about this 100 bucket limit. I get confused about what > is a bucket and what is an object in S3. Anyway, there are hundreds of > files in S3 now. The cron job stores them in the directory tree path. > (Not really true since the 'path' is just a key to the object in S3.) > > The bash backup script is a cron job that runs every few minutes. The > perl rewritemap program is run from Apache on the app role. > > The only problem I have with this setup is that some users post large > numbers of files and if the cron job is still running when it is time > to call it again, I receive error messages. > > As a first refactor, I modified the cron script to delete all local > files after they are uploaded to S3. As a second refactor, I need to > modify the cron script to exit the new job if the script is still > running, but have not done this yet. > > So now, all user uploaded files are served from S3 via an Apache > rewritemapping program in perl, there are several hundred files in S3, > the files are not open to the public for access, and I have not had > any user complaints about slow access time. I use one of the websites > and have not noticed any significant latency. > > Maybe this approach will be helpful. > > > --~--~---------~--~----~------------~-------~--~----~ You received this message because you are subscribed to the Google Groups "scalr-discuss" group. To post to this group, send email to [email protected] To unsubscribe from this group, send email to [email protected] For more options, visit this group at http://groups.google.com/group/scalr-discuss?hl=en -~----------~----~----~----~------~----~------~--~---
