total share count impacting due to Cache configured on server

Hello,

do you have some real url′s for testing?

Its important that you purge your cache after creating a new page.

Hi Rene

Thanks for quick response

Article
https://www.rd.com/culture/finish-this-sentence-the-title-of-my-autobiography-would-be/

showing total no of count : 8.5k

Actual total Share count for above url is
Likes: 196
Shares: 25
Comments: 26
Total: 247

Other existing articles
https://www.rd.com/culture/most-stolen-social-security-number/ showing 8.5k to users
actual total share for articles is :
Likes: 149
Shares: 29
Comments: 2
Total: 180

So let suggest

can you purge mashare cache and enable debug mode? i will check it than again.
go to mashshare->settings->debug

Hi Rene

I enabled debug mode and purge mash-share cache options in plugin settings

https://www.rd.com/culture/most-stolen-social-security-number/
https://www.rd.com/culture/finish-this-sentence-the-title-of-my-autobiography-would-be/

So Please have a look when you got in

-Thanks

Rene,

Every time we publish a new post we will need to clear the mashshare cache?

I checked it and Mashshare is working fine. Its internally counting the share count correct for the mentioned posts. The problem is that your memcached or probably does not expire or the edgecast ttl is too high. Please ask your developer how to expire the memcache cache every x hours so that the count will be returned correct.

Read here for more: https://stackoverflow.com/questions/1418324/memcache-maximum-key-expiration-time

What does the custom fields contain for the mentioned pages?
Check out the screenshot: https://take.ms/12yrC

I will extend Mashshare to a ajax based count function with one of the next updates to get rid of such caching issues but for know its important that you find a way to adjust the expiration time of your memcached server.

@mikel King No, not necessary. I am trying to solve a specific problem for @rpandit which is probably not related to your server constellation.

If you have a staging platform I’d love to assist you with this. You can reach me directly at info [at] mashshare.net.

I see that you ve got this problem not on every page:
https://www.rd.com/slideshows/freezing-and-defrosting-tips/view-all/

Hi Rene

Thanks for writing valuable support and solution about it.sure we would look into our expiration of memcached/edgecast settings on server , and also we are getting mash share custom fields data in db records

as you suggested purging mash-share cache on newly created article.

it could be possible to have some patch to plugin which purge the mash share cache at time of article published so that total counting should display looks updated to articles

Our approach to do purge cache for each article, enable a add option in Mash share settings

@rene /@Mikel King what are your thought about it

Please share your thought & suggest.

We actually have three levels of caching, varnish, memcache and VMDS (Edgecast CDN). They are all configured at varying lengths.

VDMS 1 hr
memcache 2 hrs
varnish 1 day

Rahul is suggesting that we use the publish article hook to clear the mashshare cache every time a new post is published and tie that to an optional check box on the plugin settings page.

Hey guys,

ok, purging the mashshare cache solved your problem?

I see that the count is correct now:
https://www.rd.com/culture/most-stolen-social-security-number/

When did this happened, after purging the mashshare cache or after expiration of the varnish cache? (I assume you guys all belong to rd.com?)

I still wonder where the 8.5k was coming from. Did you used the fake count…for a test?

Another question: Are you using sharedcount.com or mashengine?

Extending Mashshare with an option field option to purge the cache only for a specific post is possible but i am still not sure if this is the solution here because you say that you already have the correct entries in the custom meta field which are not returned. Purging the cache in mahshare only means that these custom meta fields will be removed entirely in db.

Here you find a function to purge the cache for a post:

function mashsb_delete_post_cache(){
    global $post;
        delete_post_meta( $post->ID, 'mashsb_timestamp' '');
        delete_post_meta( $post->ID, 'mashsb_shares', '' );
        delete_post_meta( $post->ID, 'mashsb_jsonshares', '' );
}

Hi Rene

We are using “sharedcount.com” with end point plus.sharedcount.com and a we have not enabled any fake count settings to plugin and total share count looks fine when we enabled debug mode or delete cache of mash share.

and i had disabled debug mode due to some page load performance and again check mash custom fields value in screen https://snag.gy/foown.jpg
yeah we have got function in plugin to purge cache for posts.

-Thanks

Hi,

please switch to mashengine and check if its throwing 8.5k again. Than we know where the wrong initial share count is coming from initially.

I have the feeling that sharedcount is delivering the wrong count sometimes. I had similiar issues in the past, so i created the mashengine.

Hi Rene

Sure we are checking it , also we have implement purge cache for each post, so we r going through checks for total share count how it working after delete cache for each post, i am also doing check for mashengine on test server & on production server too

-Thanks

Great, let me know how is it going and if i can be more helpful for you.