0

I am trying to solve this logic problem but my current MySQL schema is giving me a very poor performance with a large dataset. I have a table which contains almost half a billion telephone records. They are called the Do Not Call list or simply the DNC. When I start my campaign to dial numbers, I have to check against this list to see if my phone numbers are there in the list or not. If yes, then scrub from the list and do not dial that number.

Here is my dnc table,

CREATE TABLE `dnc_contacts` ( `id` BIGINT(20) NOT NULL AUTO_INCREMENT, `group_id` INT(11) NOT NULL, `contact` VARCHAR(45) NOT NULL, `status` TINYINT(1) NOT NULL DEFAULT '1', `created_date` TIMESTAMP NULL DEFAULT CURRENT_TIMESTAMP, `admin` TINYINT(1) NOT NULL DEFAULT '0', `updated_date` TIMESTAMP NULL DEFAULT CURRENT_TIMESTAMP ON UPDATE CURRENT_TIMESTAMP, PRIMARY KEY (`id`,`group_id`), KEY `fk_dnc_contact_dnc_group1_idx` (`group_id`), KEY `idx-group_id-contact` (`group_id`,`contact`), CONSTRAINT `dnc_contacts_copy_ibfk_1` FOREIGN KEY (`group_id`) REFERENCES `dnc_groups` (`id`) ON DELETE CASCADE ON UPDATE CASCADE ) ENGINE=InnoDB DEFAULT CHARSET=latin1;

The query I am using to compare the dnc table is;

UPDATE contacts c LEFT JOIN dnc_contacts d ON d.contact = c.contact AND d.group_id = '9' SET campaign_id = '525', call_status = 1, active_call = 0, dnc_contact = if(d.contact IS NULL, 0, 1) WHERE phonebook_id IN (920)

and my contacts table looks like;

CREATE TABLE `contacts` ( `id` bigint(20) NOT NULL AUTO_INCREMENT, `phonebook_id` int(11) NOT NULL, `campaign_id` int(11) DEFAULT NULL, `contact` varchar(45) NOT NULL, `status` tinyint(1) unsigned NOT NULL DEFAULT '1', `admin` tinyint(1) unsigned NOT NULL DEFAULT '0', `call_status` tinyint(1) unsigned NOT NULL DEFAULT '1', `active_call` tinyint(1) unsigned NOT NULL DEFAULT '0', `test_call` tinyint(1) unsigned NOT NULL DEFAULT '0', `dnc_contact` tinyint(1) unsigned NOT NULL DEFAULT '0', `v_node_name` varchar(45) DEFAULT NULL, `updated_date` timestamp NOT NULL DEFAULT CURRENT_TIMESTAMP ON UPDATE CURRENT_TIMESTAMP, `created_date` timestamp NULL DEFAULT NULL, PRIMARY KEY (`id`,`phonebook_id`), KEY `fk_contact_phonebook1_idx` (`phonebook_id`), KEY `idx-phonebook_id-contact` (`id`,`contact`) ) ENGINE=InnoDB DEFAULT CHARSET=latin1;

Running this query for small datasets with 100K to 200K contacts works fine, but with large dataset with more than 300 million records kills the performance. Can't think of a better way to achieve the goal.

Appreciate any advises to improve the logic/schema/query ..

Thanks.

0

Isn't contact unique in both tables? If so, get rid if id and have PRIMARY KEY(contact). There are multiple reasons why this is likely to speed things up.

active_call and call_status sound like things that are changed often? That interferes with the setting of the dnc_contact, especially when done in large chunks.

I can think of no good reason to even have dnc_contact in contacts. It is redundant information. That is, when you are about to make a call, do the JOIN to verify that the phone number is not in dnc_contact.

If you really need to do the UPDATE, here are tips on how to do it is a relatively non-invasive way -- bu chunking.

  • What I agree - it is for so big dataset, possible to use stored procedures for update by chunks, this is good solution – a_vlad Oct 25 '16 at 5:11
  • No the contact is not unique. I can have the same contact but under a different group or a phoneook. It gets even messier when there are multiple phonebooks. I am willing to change the schema but can't think of anything better. active_call and call_status keep on changing as the numbers are dialed. – NBhatti Oct 25 '16 at 11:30
  • It may be better to move those flags to another, parallel, table (vertical partitioning). There are too many unknowns for me to be sure of improvement. Chunking, on the other hand, should help. – Rick James Oct 25 '16 at 17:28

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.