Use complete spike rush redis watch function (rpm)

redis use watch spike rush to complete the function:

Redis completed using two key functions rush spike, mywatchkey for storing the number of users buy and store mywatchlist buy list.

Its advantages are as follows:

1. First choose to buy fast memory database.

2. Fast concurrent nature no not a problem.

3. Use the pessimistic locking, quickly increasing system resources.

4. stronger than the queue, the queue will instantly bursting your memory database resources.

5. Use optimistic locking, achieve comprehensive needs.

 

I think the following code is certainly what you want.

 

1. <?php  
2. header("content-type:text/html;charset=utf-8");  
3. $redis = new redis();  
4. $result = $redis->connect('10.10.10.119', 6379);  
5. $mywatchkey = $redis->get("mywatchkey");  
6. rob_total = 100 $;    // number of panic buying  
7. if($mywatchkey<$rob_total){  
8.     $redis->watch("mywatchkey");  
9.     $redis->multi();  
10.       
11.     // set the delay, convenient test results.  
12.     sleep(5);  
13.     // insert snapped data  
14.     $redis->hSet("mywatchlist","user_id_".mt_rand(1, 9999),time());  
15.     $redis->set("mywatchkey",$mywatchkey+1);  
16.     $rob_result = $redis->exec();  
17.     if($rob_result){  
18.         $mywatchlist = $redis->hGetAll("mywatchlist");  
19.         echo  "buy success <br/>!";  
20.         echo  "remaining number:" (. $ rob_total- $-mywatchkey. 1). "a";  
21.         echo  "User List: <pre>";  
22.         var_dump($mywatchlist);  
23.     }else{  
24.         echo  "I'm feeling good, then buy!"; Exit;  
25.     }  
26. }  
27. ?>