ddtechsolutions

The certificate works just fine, I have set it up the same as the other websites. However, one thing where I noticed it was different was in the URL, one had www and the other was missing the www part in the URL when I compared to sites where it was working well. So I matched that and now it works. However, the WP-Piwik confirmation _was_ working without changes made on the server side, which would imply something to do with the WP-Piwik plugin changing that could have caused the issue.

I have been having this very same issue. Is this a bug in this version? To my knowledge, it stopped working soon after 1.0.7 was installed. I’m not sure the exact date, I’m only noticing this on one website which has the same settings in WP-Piwik & connecting to the same Piwik back-end server, and the traffic on this particular site is very low, so I can’t quite tell when it broke.

Hi there,

Sorry for the delay. Had to finish some other projects. Back on to this again for the moment.

So you are correct, when it’s set to force SSL in the Piwik configuration, the final URL then becomes https://www.dd…ns.com/analytics/ (with the s in https instead of just http).

However, if I change the URL in the WP-Piwik plugin to include that instead, it still fails and never successfully connects.

I checked the error logs and there’s only one for the Piwik and it was an error around accessing the database, but it was several hours before the test I just did so it isn’t related. I checked the access logs as well and it’s all successful 302 and 200 codes all along the file. The ones which seem to align the time of the test appear to be 200 codes. As far as the server is concerned, everything looks fine from what I can tell in the logs. It’s being accessed successfully per the log entries.

I ran the test and got the following output as well, I hope this helps narrow it down.

————

`WP-Piwik 1.0.2
Mode: http

Test 1/3: global.getPiwikVersion

bool(false)

array(2) {
[0]=>
bool(false)
[1]=>
string(159) “https://www.dd…ns.com/analytics/?module=API&method=API.getBulkRequest&format=php&urls%5B0%5D=method%3DAPI.getPiwikVersion%26idSite%3Dn%2Fa&token_auth=…”
}

Test 2/3: SitesManager.getSitesWithAtLeastViewAccess

bool(false)

array(2) {
[0]=>
bool(false)
[1]=>
string(236) “https://www.dd…ns.com/analytics/?module=API&method=API.getBulkRequest&format=php&urls%5B0%5D=method%3DAPI.getPiwikVersion%26idSite%3Dn%2Fa&urls%5B1%5D=method%3DSitesManager.getSitesWithAtLeastViewAccess%26idSite%3Dn%2Fa&token_auth=…”
}

Test 3/3: SitesManager.getSitesIdFromSiteUrl

bool(false)

array(2) {
[0]=>
bool(false)
[1]=>
string(348) “https://www.dd…ns.com/analytics/?module=API&method=API.getBulkRequest&format=php&urls%5B0%5D=method%3DAPI.getPiwikVersion%26idSite%3Dn%2Fa&urls%5B1%5D=method%3DSitesManager.getSitesWithAtLeastViewAccess%26idSite%3Dn%2Fa&urls%5B2%5D=method%3DSitesManager.getSitesIdFromSiteUrl%26idSite%3Dn%2Fa%26url%3Dhttp%3A%2F%2Fwww.gl…al.com&token_auth=…”
}

————

Not sure if it matters, but I noticed the “method” had a value of “http” instead of “https”… does that matter or is that expected?

Also, in case it wasn’t obvious, the parts that have a “…” in the middle of a domain name is just me protecting them. Isn’t actually what’s in there. haha.

Thank you again for any help you can provide.

— Dustin

Ok, I took a bit of a closer look at this. I have found/confirmed a couple of things the last couple of days.

I followed the redirects to the ultimate destination from the URL I had in the WP-Piwik plugin, using this tool: https://redirectcheck.com/index.php

I found that the ultimate URL is what is number three above which I copied below for reference too.
(3) https://www.dd…ns.com/analytics/)

But even with that URL in the URL bar for WP-Piwik, connecting over SSL does not work. So then I changed the URL to read “https://www.dd…ns.com/analytics/” (note the HTTPS) and it still did not connect. But when I tried the diagnostic test again, it didn’t even give an HTTP response, so I can’t say why that is when I can connect using that URL no problem to the Piwik app via my web browser.

So then I tried connecting again with the HTTPS URL but with the “Disable SSL peer verification” but nothing changed, still unable to connect.

For what it’s worth, the method I am using to implement the HTTPS is the instructions found on this URL: https://piwik.org/faq/how-to/#faq_91

Am I supposed to be doing something different for the plugin to recognize it’s connecting via SSL instead of non-SSL? For all intents and purposes, the Piwik installation and consuming it over SSL works perfectly fine and as expected according to the Piwik documentation. Nothing _seems_ wrong on the server end. This makes me wonder if it’s a configuration problem with my WP-Piwik plugin or if it’s not recognizing something it needs, that maybe I can help expose to it.

Any other suggestions would be great. Thank you!

PS – Yes, the WordPress and Apache are running on the same server, however not all my websites are on the same server, so using the PHP connection would not be an option for all cases, unfortunately.

Hi braekling,

I have three different URL’s (non-SSL) which can access the Piwik web app. They are:

1) https://analytics.dd…ns.com/
2) https://dd…ns.com/analytics/
3) https://www.dd…ns.com/analytics/

Number one forwards to number three. Number two also forwards to number 3. I can reproduce this in my web browser. Number three is the ultimate destination.

I caught that the URL I had in WP-Piwik was number two, so following your advice I changed it to number three. However it still occurs.

Any other possible root causes you can think of? I did some research and sure enough the 302 does suggest a redirect, but I mean I just take the URL shown in the web browser, so I am not sure what other redirect occurs beyond that.

Hi there,

Here is the requested information. I really appreciate the assistance. ??

string(276) “HTTP/1.1 302 Found
Date: Wed, 03 Jun 2015 03:50:50 GMT
Server: Apache
X-Powered-By: PHP/5.4.16
Location: https://dd[…%5Dns.com/analytics/
Cache-Control: max-age=0
Expires: Wed, 03 Jun 2015 03:50:50 GMT
Content-Length: 0
Content-Type: text/html; charset=UTF-8

1. cURL is available.
2. allow_url_fopen is enabled.

Will the update resolve the issue or will we need to re-connect still? In other words… should I just leave it for now and update the plugin which will hopefully re-read the settings / configuration and work again? Or should I reconnect them all anyways?

I can confirm this has affected all my websites as well (the plugin’s auto-update). Settings are lost and I can reconnect no problem, but it’s at default settings again. I’m planning on just re-connecting them all (there’s about 8 websites), but please do keep this in mind when issuing a fix as it should keep the settings from the current version for those who have re-configured it, or re-read the settings from the previous version. God speed. ??

The documentation on this particular feature does leave much to be desired. I’ve just recently started using though and here’s what I can say to help: 1) The drag and drop functionality for the slider exists on the media items properties. So edit an image in the media library and you should see at the bottom a toggle for slider and the related options underneath it. If you chose the same slider as other images, then you’ll also see the other images you’ve got on the same slider at the very bottom in which you can then drag and drop to re-order.

Funny enough though, anytime I drag and drop, it says “changed saved” right away in yellow above the slider settings, but then it appears to bug out and remove the image from the slider all together. It’s an odd bug I think. So for me anyways, I know how to re-order the slides, but it also doesn’t appear to be working like it should.

Hope that offers at least some sort of assistance.

I’m having the same problem here, it creates tens of duplicate posts at times, but it ONLY seems to happen when the UNIX Cron is disabled on the Options page. If it is enabled, then it creates single posts as expected.

However that then leads me to another problem of it not running on schedule, but I’ll save that for another thread. ??

I’m having this same issue on my blog. If it helps, view the source code at https://www.ddtechsolutions.com/blog in case that shows something useful.

I’m not sure if it’s a wrong setting somewhere or if it’s a bug. ??

Yeah Rishi, this problem is definitely stemming from the 1.0.2 version from the sounds of it. Or at least many of the users on here are using that version and complaining of the problem. This version did not fix the issue, in fact from what I can tell it was the version that introduced this issue.

So what do we do then? Can we downgrade to the version that this was working in? I’m glad to know it’s only this version – I just installed it and found this issue and was like WTH ahaha, but I’m glad to see I’m not the only one and that Facebook is also working on it, presumably quickly.

Having the same problem here too, seems very unreliable right now. Seems to work for a little while (say a few hours?) then it just becomes disabled after such time has passed.
Facebook, looking for a fix here please. ??