当前位置:网站首页>10 schemes to ensure interface data security

10 schemes to ensure interface data security

2022-07-06 19:20:00 Young

In our daily development , How to ensure the security of interface data ? Personally feel , Interface data security assurance process , Mainly reflected in these aspects : One is The data transfer Safety in the process , And the data arrives at the server , how Identifying data , The last point is Security of data storage . Today, I want to talk to you about how to ensure the security of interface data 10 A plan .

 Insert picture description here

1. Data encryption , Prevent the transmission of newspaper messages .

We all know , Data in the process of network transmission , It's easy to be caught . If you are using http agreement , Because it is transmitted in clear text , Users' data can be easily obtained by others . So we need to encrypt the data .

1.1 How to encrypt data ?

Common implementation methods , That's right Key fields encryption . such as , You have a login interface , You can be right password encryption . What encryption algorithm is generally used ? Easy to use Symmetric encryption algorithm ( Such as AES) To encrypt and decrypt , Or hash algorithm processing ( Such as MD5).

What is? Symmetric encryption : An encryption algorithm that uses the same key for encryption and decryption .

 picture

Asymmetric encryption : Asymmetric encryption algorithm requires two keys ( Public key and private key ). Public key and private key exist in pairs , If you encrypt data with a public key , Only the corresponding private key can be decrypted .

A safer way , Just use Asymmetric encryption algorithm ( Such as RSA perhaps SM2), Public key encryption , Private key decryption .

 picture

If you want to encrypt all fields , It is generally recommended to use https agreement .https In fact, in the http and tcp Add a layer of encryption between SSL.

1.2 friends , Do you remember https How about the principle of ?

Interview is also often asked , as follows :

 picture

  1. Client initiated Https request , Connected to the server 443 port .
  2. The server must have a set of digital certificates ( The certificate contains a public key 、 Certification authority 、 Expiration date, etc ).
  3. The server sends its own digital certificate to the client ( The public key is in the certificate , The private key is held by the server ).
  4. After the client receives the digital certificate , Will verify the validity of the certificate . If the certificate is verified , A random symmetric key will be generated , Encrypt with the public key of the certificate .
  5. The client sends the encrypted key of the public key to the server .
  6. After the server receives the ciphertext key sent by the client , Use your previously reserved private key to decrypt it asymmetrically , After decryption, you get the client's key , Then use the client key to encrypt the returned data symmetrically , The data transmitted by maozi are ciphertext .
  7. The server returns the encrypted ciphertext to the client .
  8. When the client receives , Decrypt it symmetrically with your own key , Get the data returned by the server .

Daily business , Data transmission encryption , use https That's all right. , If the safety requirements are high , Such as login and registration , Need to transmit password , The password can be used RSA And so on , Encrypt password . If your business , The security requirements are very high , You can simulate https This process , To message , Do encryption and decryption again .

2. Data signature verification

Data message signature verification , yes Common means to ensure the safety of data transmission , It can ensure that the data is in No tampering during transmission . I used to do it Enterprise transfer system , Just use the additional signature to verify the signature .

2.1 What is countersigning and signature verification ?

  • Data signature : use Hash Algorithm ( Such as MD5, perhaps SHA-256) Generate message summary from original request parameters , Then encrypt the digest with the private key , Get the digital signature corresponding to this message sign( This process is The signature of the ). Generally speaking , The requesting party will send Digital signature and original message Send it to the receiver .

 picture

  • attestation : The receiver gets the original message and digital signature (sign) after , use The same Hash Algorithm ( For example, they all use MD5) Generate a summary from the message A. in addition , Decrypt the digital signature with the public key provided by the other party , Get a summary B, contrast A and B Are they the same? , You can know whether the message has been tampered with .

 picture

Actually The signature of the , I understand , The request parameters , According to certain rules , utilize hash Algorithm + The encryption algorithm generates a The only label sign. If the signature is checked , That is to treat the request parameters according to the same rules , Then use the same hash Algorithm , And the corresponding key decryption processing , To compare whether this signature is consistent .

Another example , Some partners achieve this , Set all non empty parameters ( Contains a package AccessKey, Unique developer ID ) In ascending order , Then splice another one SecretKey( This is only used for local encryption , Not participating in network transmission , It is only used as a signature inside ), Get one stringSignTemp Value , Last use MD5 operation , obtain sign.

After the server receives the message , Can check , Only by having a legal identity AccessKey And signature Sign correct , Only then . This solves the problem of authentication and parameter tampering , If the request parameters are hijacked , Because the hijacker can't get SecretKey( For local encryption only , Not participating in network transmission ), He can't forge a legal request

2.2 With https Wait for encrypted data , Why is it necessary to add a signature for verification

Some friends may have questions , Signature verification is mainly to prevent data from being tampered with during transmission , If it's all used https The next protocol encrypts the data , Why can it be tampered with ? Why is it necessary to add a signature for verification ?

The data was encrypted during transmission , Theoretically , Even if caught , And the data won't be tampered with . however https It's not absolutely safe Of course . You can read this article : terrible , original HTTPS It's no use . There's another point :https The encrypted part is only on the Internet , Then there are many services that jump to each other on the intranet , Countersignature can also be in This is guaranteed not to be tampered with by intermediaries , Therefore, the interface development with high security requirements for general transfer , Need to be Additional signature

3.token Authorization and authentication mechanism

Daily development , Our website or APP, Are all need The user login Of . So if it is Non login interface , How to ensure safety , How to confirm the user's identity ? have access to token Authorization mechanism .

3.1 token Authorized authentication scheme

token Authorized authentication scheme : The user enters the user name and password on the client , Click to log in , The server will verify the password successfully , A unique value will be returned to the client token, And will token In the form of key value pairs stored in the cache ( It's usually Redis) in . The subsequent client should take this with it for all operations requiring authorization module token, After the server receives the request , to token verification , If token There is , It shows that it is a legal request .

token The login authorization flow chart is as follows :

 picture

  1. The user enters the user name and password , Initiate a login request
  2. The server verifies the password , If the verification passes , Generate a globally unique token.
  3. take token Stored in redis in , among key yes token,value yes userId Or user information , Set an expiration time .
  4. Put this token Return to the client
  5. When the user initiates other business requests , You need to bring this token
  6. Background services will uniformly intercept interface requests , Conduct token Validation , And get user information from it , For subsequent business logic . If token non-existent , Indicates that the request is invalid .

3.2 How to ensure token The safety of the ?token Hijacked ?

How can we guarantee token What about the safety of the company ?

for instance , If I get token, Is it possible to call any interface on the server side ? We can consider from these aspects :

  • token Set a reasonable period of validity
  • Use https agreement
  • token You can encrypt again
  • If you are accessing sensitive information , Simply add token It's not enough. , Usually, the white list will be configured

Speaking of token, Some friends may think of jwt, namely (JSON Web Token), In fact, it is token A kind of . Interested friends can go to know about it .

4. Time stamp timestamp Timeout mechanism

Data is easy to capture , Suppose we use https And countersignature , Even if the middleman catches the data message , It can't see the real data . But some lawbreakers , He doesn't care about the real data , Instead, you get the captured packets directly , Make a malicious request ( such as DOS attack ), To bring down your system .

We can introduce Timestamp timeout mechanism , To ensure interface security . Namely : Every time a user requests it, he will bring a timestamp of the current time timestamp, The server receives timestamp after , Decrypt , After passing the examination , Compare with the current time of the server , If the time difference is greater than a certain time ( such as 3 minute ), The request is considered invalid .

5.timestamp+nonce The scheme prevents replay attacks

The timestamp timeout mechanism is also vulnerable , If it's in Within the time difference , Replay attacks by hackers , That won't work . have access to timestamp+nonce programme .

nonce Refers to a unique random string , Used to identify each signed request . We can send every request nonce Parameters are stored in a “set aggregate ” in , Or maybe json Format is stored in a database or cache . Each processing HTTP When asked , First of all, judge the nonce Whether the parameter is in the “ aggregate ” in , If there is an illegal request .

However, for servers , Keep forever nonce The price is very high . Can combine timestamp To optimize . because timstamp Parameter for more than 3min Request , It's considered illegal to ask , So we just need to store 3min Of nonce Parametric “ aggregate ” that will do .

6. Current limiting mechanism

If users are real users , He maliciously calls the interface frequently , Want to bring down your system ? In this case, it is necessary to access current limiting .

There are token bucket and leaky bucket algorithms in common use . You can read my article , For the interview :4 Explanation of a classical current limiting algorithm

have access to Guava Of RateLimiter Stand alone current limiting , You can also use Redis Distributed current limitation , You can also use Alibaba open source components sentinel Current limiting . for instance , How many requests can you accept in a minute .

7. Blacklist mechanism

If malicious requests from real users are found , You can set up a blacklist mechanism , Black the user . General situation , There will be some competitors , Or users who are not bad intentions , Want to do your system . therefore , In order to ensure safety , Generally, our business system , There needs to be a blacklist mechanism . For blacklist initiated requests , Just return the error code directly .

8. White list mechanism

With the blacklist mechanism , You can also set up a white name single machine system . The enterprise transfer system I used to be responsible for , If an outside merchant wants to access our system , You need to apply for the online white list in advance . At that time, O & M will apply for IP Internet white list , Only the requests in the white list , To access our transfer system .

9. Data desensitization mask

For passwords , Or cell phone number 、 Sensitive information such as ID card , Generally, desensitization mask is required to display , If it's a password , It also needs to be encrypted and saved to the database .

For mobile phone numbers 、 ID card information , Daily development , When checking logs , What you see should be masked . The purpose is to try not to disclose these user information , Although only development and operation and maintenance can read the log , But it still needs to be prevented , Do masking .

Save the password to the database , We definitely can't directly save in plaintext . The simplest also needs MD5 Deal with it and save it ,Spring Security Medium BCryptPasswordEncoder It's fine too , Its bottom layer is SHA-256 + Random salt + secret key Encrypt password , and SHA and MD The series is the same , All are hash Algorithm of abstract class .

10. Check the validity of data parameters .

Security assurance of interface data , We also need our system , There is a data validity check , In a nutshell Parameter checking , For example, the length of ID card , Length of mobile phone number , Is it a number, etc .

summary

This paper introduces 10 A scheme to ensure the security of interface data . friends , If there are other plans , You can comment in the message area , Exchange and study together .

原网站

版权声明
本文为[Young]所创,转载请带上原文链接,感谢
https://yzsam.com/2022/187/202207061122470553.html