This is what I've been using for a reverse proxy config:
# -- BEGINING reverse proxy
ProxyTimeout 60
<Location /xyz123/>
Order Allow,Deny
Allow From all
ProxyPass
http://999.999.999.999/xyz123/
ProxyPassReverse
http://999.999.999.999/xyz123/
</Location>
# -- END reverse proxy
With some, I limit access to just internal ip's, but this is the gist of it. Thank you for passing along the link. It looks like we're using the same setup.
From: Nathan Andelin <nandelin@xxxxxxxxx>
To:Web Enabling the AS400 / iSeries <web400@xxxxxxxxxxxx>
Date: 11/3/2010 10:53 PM
Subject: Re: [WEB400] Webserving pdfs via reverse proxy, QNTC, NFS
The following is a pretty good tip for setting up an Apache reverse proxy, how
to do it, and why. Thanks to Giovanni Perotti.
http://www.mcpressonline.com/tips-&-techniques/internet/techtip-need-a-safe-ibm-i-internet-connection.html
-Nathan
----- Original Message ----
From: Jeffrey Day <Jeff.Day@xxxxxxxx>
To: web400@xxxxxxxxxxxx
Sent: Wed, November 3, 2010 7:16:42 PM
Subject: Re: [WEB400] Webserving pdfs via reverse proxy, QNTC, NFS
Hi Nathan and Scott,
Thank you for responding, I will be trying the NFS route.
The proxy is running in apache on the 400, and is our main web server. As to
why the pdf's aren't hosted on the 400? They initially were, but now there are
roughly 900,000 of them, they took up too much space, so we moved them to a
cheaper storage box. One reason for the performance hit, is that the pdfs are
grouped (think separate directories) and when a visitor searches for pdfs, I
first verify that the pdf exists (to avoid dead links) before writing the link
for it to the screen.
Up to now, I have been using GETURL to check the existence of pdfs, but I
suspect I can improve performance, by cutting out the web layer, configuring NFS
and using CHKIFSOBJ to verify files.
Thanks,
Jeff
As an Amazon Associate we earn from qualifying purchases.