All locations active · 99.99% uptime
Network Technology

Subnet Diversity and Proxy Pool Resilience

When a proxy provider says "100,000 IPs", the first question that comes to mind is this: how many different blocks do those IPs come from? Because blocking decisions are most often applied not to individual IPs but at the level of IP blocks themselves. If all 100,000 addresses come from a few hundred blocks, your real resilience is not 100,000 but a few hundred.

In this article we look at the concept of the subnet, how block-level blocking works, and how to measure the diversity of a pool.

Subnets and CIDR Notation

IPv4 addresses are 32 bits, and the /n notation states how many of the leading bits make up the "network" part. The remaining bits enumerate the addresses within that network:

NotationNumber of addressesTypical use
/321A single address — dedicated proxy
/298A small corporate allocation
/24256The classic "class C"; the unit of blocking
/221.024A mid-sized hosting block
/1665.536A large provider allocation

When the industry talks about "C-class diversity", what is meant is how many different /24 blocks the pool is spread across. This is the practical measure of resilience to blocking.

FIGUREThe block structure of an IPv4 address
ANATOMY185.93.68.14/24 → network: 185.93.68.0 · range: .0 – .255185.93.68Network part — the block identity; blocking is applied at this level.14Host part — the sequence number within the block/24The first 24 bits are the network, the remaining 8 bits the host (256 addresses)

When a site decides to "block the 185.93.68.0/24 block", all 256 addresses in that block lose access at the same moment.

Why Is Blocking Done at Block Level?

Seen from the defensive side the logic is simple: blocking abusive IPs one by one is both slow and easy to get around, since the attacker just moves to the next address. Blocking the block covers all the neighbours in a single move.

FIGUREThe block distribution of two different pools
DISTRIBUTIONBlock ABlock BBlock CBlock DBlock EBlock FConcentrated pool420380310000Dispersed pool9588102918497The same total number of IPs, very different resilience

In a concentrated pool, if a single block is blocked a third of your capacity is gone. In a dispersed pool the same event affects only a sixth of it.

Measuring the Diversity of Your Pool

If you have a list of IPs, calculating diversity is simple: take the first three octets of each address and count the unique ones.

FIGUREFinding the number of unique /24 blocks in a pool
Terminal / Python01# Bash — count the unique /24 blocks in a proxy list02cut -d: -f1 proxyler.txt | cut -d. -f1-3 | sort -u | wc -l0304# Python — distribution table05from collections import Counter06ips = [l.split(":")[0] for l in open("proxyler.txt") if l.strip()]07blocks = Counter(".".join(ip.split(".")[:3]) for ip in ips)08print("total IPs :", len(ips))09print("unique /24 :", len(blocks))10print("avg per block:", round(len(ips) / max(1, len(blocks)), 1))11print("densest 5 blocks:", blocks.most_common(5))

The lower the average number of IPs per block, the more resilient the pool. Having all 256 addresses come from the same block is, in practice, like owning a single IP.

Practical threshold

An average per block below 10 IPs counts as a good distribution. Above 50, your pool has become concentrated and a single blocking decision creates a large loss.

The Situation Is Different in Residential Pools

Residential pools have high diversity by their very nature, because the IPs come from real home subscribers and are spread across thousands of different ISP blocks. Here the real issue is not the number of blocks but how large the pool genuinely is and how often you get the same IP again.

A simple test: send 200 requests without using the same session key and count how many different exit IPs you see. If the repeat rate is high, the pool is not as large as claimed. Our residential proxy article explains the pool architecture in detail.

The Concept of Diversity Changes in IPv6

In IPv6 the address space is so large that a single customer is usually allocated an entire /64 block — which means many times more addresses than the whole of IPv4. But this abundance is misleading:

FIGUREDiversity logic in IPv4 and IPv6
COMPARISONIPv4IPv6Blocking unit/24 block/64 or /48 blockAddress abundanceScarce — expensivePlentiful — cheapPer customerA few addressesBillions of addressesMeasure of diversityHow many different /24sHow many different /64sRotation valueHighLow within a blockTarget supportUniversalPartial

In IPv6, changing addresses within the same /64 means "the same source" to most defence systems. Real diversity comes from using different /64 blocks.

For the cost advantage and the limits of IPv6, see our IPv4–IPv6 comparison and IPv6 proxy page.

Questions to Ask When Buying

  • How many different /24 blocks is the pool spread across? If no number is given, ask for a small sample and measure it yourself.
  • How many different ASNs are used? A single ASN means a single point of failure.
  • How many countries are the blocks distributed across? Critical if you are going to do geographic targeting.
  • How many customers is the same block shared with? In shared pools your neighbours' behaviour affects you.
  • Are blocked blocks replaced? The provider's replacement policy determines your resilience.

Managing Distribution in Your Own Pool

If you buy IPs from more than one provider, add block awareness to your selection logic: spread consecutive requests across different blocks as much as possible. This makes it harder for the target to see the signal of "heavy traffic from the same neighbours".

FIGUREBlock-aware selection
CHOICE6 blocksbalanced distribution1Consecutive requests from different blocksneighbour density drops2If a block is blocked, quarantine all of itdo not try them one by one3A concurrency limit per blockprevents concentration

When a block is blocked, quarantining all the addresses in it together is far more efficient than spending time and quota trying them one by one.

Summary

Pool size on its own is not an indicator of quality; the real determinant is the distribution across blocks and ASNs. Because blocking decisions are most often applied at the /24 level, concentrated pools are fragile. You can measure the list you have with a couple of commands and increase resilience by keeping the average number of IPs per block low. You can verify the liveness and exit locations of your addresses with the proxy checker tool in bulk.

Frequently Asked Questions

01How many addresses does a /24 block contain?

It contains 256 addresses (.0 through .255). Two of these are usually reserved as the network and broadcast addresses, so the number of usable addresses is 254.

02How many different blocks are enough?

It depends on your work. For low-volume work 5–10 blocks are enough, while for data collection at scale a good rule is to aim for a distribution that does not exceed 10 IPs per block.

03Should I ask about subnet diversity in a residential pool?

Rather than asking directly, go for a practical test: send a few hundred requests without using a session key and measure how many different exit IPs and how many different blocks you see.

04I have a large number of IPv6 addresses — does that solve my rotation problem?

No. Changing addresses within the same /64 is treated as a single source by most defence systems. Real diversity requires different /64 or /48 blocks.

05How do I know a block has been blocked?

If several addresses in the same block receive a 403 or a CAPTCHA at the same time on the same target, that is a strong sign. In that case, quarantine the whole block instead of trying addresses one by one.

Related Articles and Pages

NEXT STEP

Strengthen your proxy setup today.

Get started in minutes with a paid plan, or try our free proxy list first.

FREEPROXY.TR

Looking for a free proxy? You're in the right place

A complete proxy platform where you can browse up-to-date free proxy addresses, compare HTTP and SOCKS proxy types, and check your proxy connections with free tools.