[size=8pt](I removed the HTTP headers - how about adding -q to the php call?)[/size]
I have three sites configured, all work OK with browser-triggered archiving.
I have dug into the script a bit, and have tried to call the api directly. the result is the same. however, when I give idSite=1 instead of idSite=all, I get this:
<?xml version="1.0" encoding="utf-8" ?>
<result>
<error message="You can't access this resource as it requires a 'view' access for the website id = 1." />
[size=8pt](there’s no closing tag, btw. not so good)[/size]
This is odd, the admin user of course has all privileges. Or is it confused because “view” is implicit in “admin”?
which version of piwik are you using? it looks like the API is called with a token_auth that doesn’t have privileges. Are you using the latest version of the cron file?
I’m using Piwik 0.5.4, archive.sh is 2034 bytes in 55 lines, md5 is b136e7f17e1fc33336d6eef3e93ef5b2.
Yes, it looks like missing privileges. That might also be caused by me doing something wrong when trying to call the API directly, while giving an explicit idSite parameters. Note that I do not see any error when calling archive.sh itself. Is that consistent with the “no privileges” explanation?