Jump to content

jwolfe

Approved members
  • Content Count

    12
  • Joined

  • Last visited

  1. I think that works, too. I wish I could see some official documentation about it or at least have someone on the development team confirm this is the correct way to do it.
  2. I think I have it working by passing custom variables in the data-revive-source attribute of the <ins> tag. Does any one know of a reason that I shouldn't be putting this type of data in the source field? Also, I noticed that when I have multiple ads on a page that the asyncjs.php file should only be loaded once after all the ad spots. Can someone confirm that this is the proper use? I can't find documentation anywhere.
  3. Here's a little background on how I implemented with the Javascript tag. I added a keyword value to the query string this way: document.write("&amp;keyword=KEYWORD_VALUE"); Then in the delivery options I have a Site - Variable limitation set to check if the Name Contains a value that I am targeting. How can I do this with the asynchronous tag? I tried with the Source setting for the async tag but that doesn't seem to work correctly.
  4. How do I pass keywords or other custom variables through the asynchronous tag?
  5. I wanted to update this thread. We moved off Amazon four weeks ago and now we are on a dedicated server with solid state drives. It serves about 21 million impressions per day and has not had any problems. Thanks to Snoork hosting for convincing me to make the change.
  6. I also tested a real Revive query using Apache's ab tool. This is the index.php file what did the query: <?php $db = array("host"=>"localhost","db"=>"test","user"=>"root","pass"=>"root"); $db = new PDO('mysql:dbname='.$db['db'].';host='.$db['host'],$db['user'],$db['pass']); $creative_ids = array("3","4","7","8","10","11","12","13","14","15","16","17","18","22","23","24","25","26","27","32","33","34","35","36","37","38","39","40","44","45","46","47","48","53","54","55","56","57","58","59","60","61","62","63","64","71","72","73","74","75","76","81","82","83","84","90","91
  7. Thanks for all the responses. I understand the concerns about cloud and virtualized servers. I have been testing on two AWS instances to compare Percona to the out-of-the-box MySQL. I don't see any improvements with the Percona server when it is doing writes. I see a big improvement when doing reads, but that isn't the issue. I can cache reads on the application server to solve the read issue. I tested using sysbench: sysbench --test=oltp --oltp-table-size=1000000 --mysql-db=test --mysql-user=root --mysql-password='root' --max-time=60 --oltp-read-only=off --oltp-index-updates=on --oltp-
  8. We haven't made any changes to our set up yet. We moved into a new office this week and we've spent most of our time packing and unpacking.
  9. Thanks for the suggestions. We are looking into changing our setup. I don't have a screenshot, but I did capture 'SHOW FULL PROCESSLIST' in spreadsheet form. Most of the connections have a State of 'update'. I parsed it and put it in spreadsheet form to try to see if there was one creative_id or zone_id that was more common that others. I'll spare you all 2,945 rows, so here are the first 50: Id User Host db Command Time State date creative zone count 126053658 revive ip-10-146-162-42.ec2.internal:37375 revive Query 62 query end 2014-04-04 22:00:00' '110' '14' '1' 126053706 revive
  10. Thanks. We have upgraded the database to an instance with the most memory to get more max_connections. We serve about 10,000,000 impressions per day. About once per day a log jam forms and all the queries are the above query. Would the same problem happen if I set up a local database on each server and then replicated it to a larger instance?
  11. It is only using 1 DB instance and it is not the SSD type. It is not using replication. That is something we were thinking about looking into, but not sure if that will solve this particular issue.
  12. We are hosting revive on Amazon Web Services. There are three web servers behind a load balancer all using the same RDS MySQL database. According to the database metrics, the server usually averages less than 20 connections at any instant. Then all of a sudden the number of DB connections are maxed out. Running SHOW FULL PROCESSLIST reveals that they are all running similar INSERT/UPDATE statements. INSERT INTO rv_data_bkt_m (interval_start, creative_id, zone_id, count) VALUES ('2014-04-04 23:00:00', '90', '14', '1') ON DUPLICATE KEY UPDATE count = count + 1 T
×
×
  • Create New...