We're still seeing issues with 404s on our CSS and JS files. For example, our page was looking for css_2a4acb2a9bd34c1e0e9e18ddec8dec6d_0.css. This became css_2a4acb2a9bd34c1e0e9e18ddec8dec6d_0.css?redirect_counter=6 and 404'd on us. On the server, we see 3 css files with similar names:
css_2a4acb2a9bd34c1e0e9e18ddec8dec6d_1.css
css_2a4acb2a9bd34c1e0e9e18ddec8dec6d_2.css
css_2a4acb2a9bd34c1e0e9e18ddec8dec6d_3.css
When I try to look for css_2a4acb2a9bd34c1e0e9e18ddec8dec6d_4.css, I'm redirected to css_2a4acb2a9bd34c1e0e9e18ddec8dec6d_3.css. Is it possible for us to get the same behavior when the increment is at 0? In the advagg_bundles table, the counter for the "2a4acb2a9bd34c1e0e9e18ddec8dec6d" bundle is set to 3. Hope that helps. We've also been seeing this issue on css and js files where the increment on the filename is 0.
Comment | File | Size | Author |
---|---|---|---|
#2 | advagg-1176898-1.patch | 432 bytes | mikeytown2 |
Comments
Comment #1
djbobbydrake CreditAttribution: djbobbydrake commentedAlso, we're running the 2011-05-27 dev release.
Comment #2
mikeytown2 CreditAttribution: mikeytown2 commentedMoving #1176346: Endless css load (404) to here as this is a better bug report.
I think this patch should fix it. Issue has to do with 0 as a number vs 0 as a string vs 0 as FALSE. Patch has been committed.
Comment #3
djbobbydrake CreditAttribution: djbobbydrake commentedVerifying that this patch fixed the issue. I simulated the bug conditions described above and the css was properly redirected after the first redirect counter increment. Thanks!