Php curl dns cache

Pay DNS with IPv6 and return the status code

problem

pass through CURLOPT_IPRESOLVE Control uses IPv4 to resolve domain names or IPv6 resolution,

However, when the resolution domain name does not exist IPv6, still returns 200 status code,

curl_setopt($ch, CURLOPT_NOSIGNAL, true); curl_setopt($ch, CURLOPT_IPRESOLVE, 4 == $iptype ? CURL_IPRESOLVE_V4 : CURL_IPRESOLVE_V6);

Solutions

Consider the timeout retry;

Because CURL is requested by default, the IPv6 address of the domain name resolution will be requested. If the IPv6 address request fails, the normal IPv4 address is requested after Timeout seconds, so this time we enforce the CURL request, use IPv4 addresses

The result is that the result is to use IPv4 address resolution, so it has been reported 200;
Because of curl_setopt($ch, CURLOPT_NOSIGNAL, true); The DNS analysis is not overtime limit, which will automatically parse IPv4;
Open curlopt_nosignal caused DNS restrictions

Consider DNS cache;

So find how to clean up DNS cache;

Unable to manually remove DNS cache from PHP. But you can use CURLOPT_DNS_CACHE_TIMEOUT You can set the option to 0 for the next request to force the CURL DNS cache.

Pass a long, this sets the timeout in seconds. Name resolves will be kept in memory and used for this number of seconds. Set to zero to completely disable caching, or set to -1 to make the cached entries remain forever. By default, libcurl caches this info for 60 seconds.

curl_setopt($ch, CURLOPT_DNS_CACHE_TIMEOUT, 0); // Disable DNS cache invalid curl_setopt($ch, CURLOPT_DNS_USE_GLOBAL_CACHE, false); // Disable DNS global cache is valid Also add: Disable link cache curl_setopt($ch, CURLOPT_HTTPHEADER, $header); // A group of http header fields curl_setopt($ch, CURLOPT_FRESH_CONNECT, true); // Force a new connection instead of the connection in the cache

Final result

function httpGetWebInfo($url, $iptype = 6, $resolve = [], $timeout = 120)   $data = []; $ch = curl_init(); curl_setopt($ch, CURLOPT_URL, $url); curl_setopt($ch, CURLOPT_CUSTOMREQUEST, 'GET'); // POST method submit curl_setopt($ch, CURLOPT_NOSIGNAL, true); // True, will cause DNS resolution from time to time control curl_setopt($ch, CURLOPT_CONNECTTIMEOUT_MS, 1000 * $timeout); curl_setopt($ch, CURLOPT_TIMEOUT_MS, 1000 * $timeout); curl_setopt($ch, CURLOPT_TIMEOUT, $timeout); curl_setopt($ch, CURLOPT_HEADER, false); // Do not need to output head information -i curl_setopt($ch, CURLOPT_NOBODY, true); // Do not output content curl_setopt($ch, CURLOPT_IPRESOLVE, 4 == $iptype ? CURL_IPRESOLVE_V4 : CURL_IPRESOLVE_V6); curl_setopt($ch, CURLOPT_RETURNTRANSFER, false); // Set whether to store the response result into the variable, 1 is deposited, 0 is the direct ECHO; curl_setopt($ch, CURLOPT_FOLLOWLOCATION, true); // Grab the jump - L --Location if (!empty($resolve))   // curl -L -I -6 www.a.com --resolve www.a.com:80: curl_setopt($ch, CURLOPT_RESOLVE, $resolve); curl_setopt($ch, CURLOPT_DNS_USE_GLOBAL_CACHE, false); // Disable DNS global cache key > curl_exec($ch); if (!curl_errno($ch))   $data = curl_getinfo($ch); > curl_close($ch); return $data; >

Источник

How to flush php curl’s dns cache?

Pass a long, this sets the timeout in seconds. Name resolves will be kept in memory and used for this number of seconds. Set to zero to completely disable caching, or set to -1 to make the cached entries remain forever. By default, libcurl caches this info for 60 seconds.

You can set option to 0 for next request to force curl’s DNS cache to be flushed

Dns Caching in Linux

On Linux (and probably most Unix), there is no OS-level DNS caching unless nscd is installed and running. Even then, the DNS caching feature of nscd is disabled by default at least in Debian because it’s broken. The practical upshot is that your linux system very very probably does not do any OS-level DNS caching.

You could implement your own cache in your application (like they did for Squid, according to diegows’s comment), but I would recommend against it. It’s a lot of work, it’s easy to get it wrong (nscd got it wrong. ), it likely won’t be as easily tunable as a dedicated DNS cache, and it duplicates functionality that already exists outside your application.

If an end user using your software needs to have DNS caching because the DNS query load is large enough to be a problem or the RTT to the external DNS server is long enough to be a problem, they can install a caching DNS server such as Unbound on the same machine as your application, configured to cache responses and forward misses to the regular DNS resolvers.

Caching DNS using getaddrinfo

In general, no, the glibc resolver does not cache getaddrinfo results.

On systems running nscd , some names may be cached. (At least, gethostbyname may go through cache; I’m not sure about getaddrinfo .) Restart the daemon to clear the cache.

It is usually expected that the upstream (presumably recursive) DNS resolver is performing caching. In most situations, that is outside of your control.

How does PHP cURL DNS caching works?

I don’t know how your CentOS and on Debian is configured to handle DNS caching, generally all operating systems are reading the time to refresh the DNS cache from your domain SOA records.

Here is how you can check any domain SOA value:
https://mxtoolbox.com/SOALookup.aspx

Here is the situation with PHP cURL:

DNS cache is implemented on cURL on Jan 2002. Read more about it here:

You can turn DNS cache for cURL ON or OFF with

CURLOPT_DNS_USE_GLOBAL_CACHE

TRUE to use a global DNS cache. This option is not thread-safe and is enabled by default.

How to clear DNS cache in google chrome

What happens if you open developer console F12 and then hold down on the refresh button and then select empty cache and hard reload?

Take a look at this gif for an example.

Where is the DNS cache stored for Ubuntu?

Firstly, you’re looking for the DNS cache maintained by curl (and not a system wide global cache) because unless you’re running a caching resolver (LDNS server), the responsibility to cache DNS records falls on the applications.

Secondly, you can use /etc/hosts to store mappings. If you’re resolver library isn’t preferring /etc/hosts over a DNS lookup, you need to change the order of lookup in /etc/host.conf. You should see a line like:

hosts here tells the gethostbyname , etc. resolver library utilities to look at the /etc/hosts file first.

How to make res_query work with dns caching?

Both dig and res_nquery() will just send the queries to whatever’s in your /etc/resolv.conf file, i.e. your dnsmasq instance. Both will also by default set the RD bit to request recursion.

dnsmasq will then either serve the answers from cache, or go fetch them if the TTL has expired. However to my knowledge there is nothing in the DNS protocol itself (and therefore nothing in the res_nquery settings) that a client can set that will influence that behaviour

Источник

Читайте также: Java число в byte

Php curl dns cache