Question 1

What causes blockchain congestion?

Accepted Answer

TL;DR: Blockchain congestion happens when more transactions are sent to a network than fit in the next block. Every chain has a cap on how much computation or data it can process per block. When demand goes past that cap, a backlog builds in the mempool, the waiting area for unconfirmed transactions. Congestion means higher fees as users compete for limited block space, slower confirmations, and a worse user experience. The root causes are fixed block capacity, demand spikes from popular events, fee market dynamics, and design choices that put security and decentralization ahead of raw throughput. The Simple Explanation A blockchain processes transactions in batches called blocks. Each block has a capacity limit: Ethereum: a gas limit, currently around 30 million gas per blockBitcoin: a weight limit of 4 million weight units per blockSolana: a mix of compute units and account locks per slot Whatever the mechanism, every chain has a ceiling on how much work fits in a block. When fewer transactions arrive than a block can hold, everything runs smoothly. Transactions land in the next block, fees are low, and confirmation is fast. When more transactions arrive than a block can hold, a queue forms. That queue is the mempool: a holding area on each node where valid but unconfirmed transactions wait for a future block. The mempool isn't first-come, first-served. It's a priority queue sorted by fee. Block producers (miners or validators) pick the transactions that pay the most, because that earns them the most. When the mempool is full, users who want fast confirmation have to outbid everyone else with a higher fee. This bidding war is what makes gas prices spike during congestion. Users who can't or won't pay the higher fees see their transactions stuck in the mempool for minutes, hours, or even days. Demand Spikes The most visible cause of congestion is a sudden surge in demand for block space. Specific events usually trigger these spikes: NFT mints: a popular mint can create tens of thousands of transactions within minutes as collectors rush to grab limited tokens.Launches and airdrops: a new DeFi protocol or airdrop claim brings a flood of users to new smart contracts at once.Market crashes: a sharp drop sets off a wave of liquidations, closed positions, and panic selling as DeFi users rush to adjust. Some of Ethereum's worst congestion came from single apps: The CryptoKitties craze in December 2017 clogged the network for days.Yuga Labs' Otherside NFT mint in May 2022 pushed gas prices past 8,000 gwei. Some users paid thousands of dollars in fees for transactions that still failed.Airdrop claims for Uniswap's UNI and Arbitrum's ARB caused congestion that lasted hours as millions of eligible wallets rushed to claim. Meme token trading causes repeated congestion on chains like Solana and Base, where one viral launch can create millions of swaps within hours. With planned events, users at least know congestion is coming. Meme token surges are unpredictable and can overwhelm a network with little warning. Fixed Block Capacity and Production Rate The deeper cause is that block capacity and block timing are fixed by the protocol, while demand swings, sometimes wildly. Ethereum produces a block every 12 seconds with a gas limit of about 30 million. Typical costs: A simple ETH transfer: 21,000 gasA Uniswap swap: 150,000-300,000 gasA complex DeFi interaction: 500,000 gas or more In ideal conditions, that works out to roughly 15-30 transactions per second, depending on how complex they are. When thousands of users try to mint an NFT, swap a token, or claim an airdrop at once, 15-30 TPS is nowhere near enough. Bitcoin has the same constraint with even tighter limits. It produces a block every 10 minutes, and each block holds about 2,000-3,000 transactions. That's roughly 5-7 TPS. When demand is high, Bitcoin's mempool can fill with hundreds of thousands of unconfirmed transactions, and low-fee ones can wait days. These limits aren't arbitrary. They protect decentralization. If blocks were bigger or faster, fewer nodes could keep up with the data, and the network would centralize around operators with expensive hardware. This tradeoff between throughput and decentralization is central to blockchain design. It's often called the blockchain trilemma: the difficulty of getting scalability, security, and decentralization all at once. MEV and Priority Gas Auctions Maximal Extractable Value (MEV) adds to congestion in a less obvious way. MEV is the profit block producers and searchers can make by reordering, inserting, or censoring transactions in a block. When a profitable MEV opportunity appears, such as an arbitrage between two DEXes, several searchers race to capture it by sending transactions with higher and higher fees. This priority gas auction eats block space and raises fees for everyone, even users whose transactions have nothing to do with it. Sandwich attacks are an MEV pattern that hits regular users directly. A searcher spots a large swap in the mempool. They place a buy order just before it (front-running) and a sell order just after it (back-running), profiting from the price move the victim's swap causes. Those extra transactions take up block space and add congestion beyond normal user demand. Layer 2 Solutions and Congestion Relief Layer 2 networks like Arbitrum, Base, Optimism, and zkSync were built partly to relieve congestion on Ethereum L1. They process transactions on a separate chain and post compressed summaries back to Ethereum, which greatly increases total throughput. A transaction that costs $5 in gas on a congested Ethereum L1 might cost $0.01 on an L2. L2s can still congest during extreme spikes, though, and they inherit some limits from the L1 they settle to. Solana handles congestion differently, with local fee markets. Instead of one global fee market where every transaction competes for the same space, Solana tries to isolate congestion to specific "hot" accounts. Transactions that touch a busy contract, like a popular DEX pool, pay higher fees. Transactions that touch unrelated contracts aren't affected. This limits the spillover from congestion spikes, though it doesn't remove them. How long does blockchain congestion last? It depends on the cause. Scheduled events like a big NFT mint or airdrop claim usually cause a sharp spike that clears within minutes to a few hours, once the rush dies down.Sustained congestion is different. When a chain runs near capacity for a long time, such as during a long bull market or a viral meme token cycle, high fees and slow confirmations can last for days. Bitcoin congestion usually clears more slowly than Ethereum's. Its 10-minute block time means the mempool drains in larger, less frequent steps. The clearest sign congestion is easing is a shrinking mempool along with falling priority fees. Congestion is closely related to why blockchains slow down in general, but it's not an outage. If a network stops producing blocks entirely, that's a chain halt, not congestion. How do block limits differ across blockchains? Every chain caps how much work fits in a block, so how much room each network has before congestion sets in varies widely. The table below shows the approximate limits that decide how fast each network backs up under load. BlockchainBlock timeCapacity metricApprox. throughputBitcoin~10 minutes4M weight units5 to 7 TPSEthereum L1~12 seconds~30M gas per block15 to 30 TPSSolana~400 ms slotCompute units and account locksThousands of TPSArbitrum / Base (L2)Sub-secondL2 gas, settled to L1Hundreds to thousands of TPS Higher-throughput chains and Layer 2s absorb demand spikes better, but no design is immune. For more on how speed and capacity relate, see throughput vs latency and what a Layer 2 blockchain is. What is the difference between congestion and a chain halt? To a user, congestion and a chain halt can both look like "the network is broken." They are different problems with different causes and fixes. During congestion, the chain still produces blocks; they're just full and expensive. During a chain halt, block production stops entirely. AspectCongestionChain haltBlock productionContinues normallyStoppedTransactionsConfirm slowly, high feesDo not confirm at allRoot causeDemand exceeds block capacityConsensus failure or client bugTypical resolutionDemand subsides, fees fallValidators coordinate a restart In short, congestion is a pricing and capacity problem, and a chain halt is an availability problem. MEV activity like priority gas auctions can make congestion worse. Knowing what MEV is helps explain why fees climb faster than user demand alone would suggest. How can developers reduce the impact of congestion? Apps can't prevent congestion, but they can stay reliable through it: Use dynamic fee estimation instead of hardcoded gas prices.Add retry logic with proper nonce management.Queue non-urgent transactions for quieter periods.For data, use push-based delivery instead of tight polling loops so ingestion keeps up as blocks grow. Knowing how RPC requests work, choosing streaming over polling, and planning for provider RPC rate limiting all lower the chance that congestion leads to lost data or failed transactions. Frequently Asked Questions Why do gas fees spike during congestion? Block space is limited, and the mempool is a fee-based priority queue. When more transactions compete than a block can hold, users bid higher fees to get included first. That bidding war pushes the prevailing gas price up until demand falls back below block capacity. Does congestion mean the blockchain is down? No. During congestion the chain keeps producing blocks and confirming transactions; they are just slower and more expensive. A network that stops producing blocks is experiencing a chain halt, which is a separate problem. Can Layer 2 networks get congested too? Yes. Layer 2s offer far more throughput than Ethereum L1, so they tolerate larger spikes, but extreme demand can still fill their capacity and raise fees. They also inherit some constraints from the Layer 1 they settle to. How can I tell if congestion is clearing? Watch the mempool size and priority fees. A shrinking backlog of unconfirmed transactions combined with falling priority fees is the clearest sign that demand has dropped back below block capacity. Why don't blockchains just make blocks bigger? Larger or faster blocks raise the hardware cost of running a node, which pushes out smaller operators and centralizes the network. Capacity limits are a deliberate tradeoff that protects decentralization, a tension often called the blockchain trilemma. How Congestion Impacts Developers Building on Quicknode Congestion affects both how developers send transactions and how they ingest data. On the sending side, transactions with fees that are too low get stuck in the mempool or dropped. Apps need solid fee estimation, retry logic, and nonce management to handle congestion well. Quicknode's Core API delivers transactions reliably with steady performance even at peak congestion. Enhanced API methods help apps estimate the right gas price for current network conditions. On the data side, congestion means more transactions per block, bigger blocks, more events, and more data. Polling-based systems struggle here. Each block holds more data while the polling interval stays the same, so a backlog builds. Quicknode Streams handles congestion well because its push-based design scales with block size. Whether a block holds 100 transactions or 1,000, Streams delivers the full data to your destination with the same guaranteed delivery and ordering. Configurable batching and compression help manage the extra data during long congestion periods. Further Reading What is a Mempool? - Quicknode GuideGuide to Efficient RPC Requests - QuicknodeGetting Started with Streams - Quicknode DocsQuicknode Core API

Question 2

Blockchain throughput vs latency

Accepted Answer

Throughput and latency are two different measurements of blockchain performance that are frequently confused. Throughput is capacity: how many transactions the network can process per second (TPS). Latency is speed: how long it takes for an individual transaction to be confirmed and finalized. A highway analogy helps clarify the distinction. Throughput is how many cars the highway can carry per hour. Latency is how long it takes one car to drive from entrance to exit. A highway can have high throughput (many cars) but high latency (slow speed due to congestion), or low throughput (few cars) but low latency (each car moves fast). The best blockchains optimize for both, but tradeoffs are real.

Question 3

Why blockchains slow down

Accepted Answer

Blockchains slow down when demand for block space exceeds the network's processing capacity. The most common causes are network congestion (too many transactions competing for limited block capacity), state bloat (growing storage requirements that increase the computational cost of processing each block), consensus overhead (communication costs that scale with validator count), and infrastructure limitations (slow nodes, high-latency RPC endpoints, and inefficient data pipelines). Slowdowns manifest as longer confirmation times, higher fees, stale data in applications, and degraded user experience. Some causes are inherent to blockchain design, while others can be mitigated with better infrastructure.

区块链	时段	容量指标	约吞吐量
Bitcoin	约10分钟	4M 重量单位	5 至 7 TPS
Ethereum L1	约12秒	每个区块约3000万gas	15 至 30 TPS
Solana	约400毫秒的时隙	计算单元和账户锁定	数千个临时保护身份（TPS）
Arbitrum /Base (L2)	亚秒	L2 气，结算至 L1	数百至数千个 TPS

方面	拥堵	链停
块生产	照常进行	停止
交易	确认过程缓慢，手续费高	完全不要确认
根本原因	需求超过区块容量	达成共识失败或客户端错误
典型分辨率	需求减弱，费用下降	验证者协调重启

想随时了解最新动态吗？

开发者工具

文档与指南

是什么导致了区块链拥堵？

简明解释

需求激增

固定区块容量和产量

MEV 与优先天然气拍卖

第二层解决方案与拥塞缓解

区块链拥堵会持续多久？

不同区块链的区块限制有何差异？

拥塞和链式停机有什么区别？

开发商如何减轻拥堵的影响？

常见问题解答

为什么在网络拥堵时 gas 费用会飙升？

网络拥堵是否意味着区块链已无法运行？

第二层网络也会出现拥塞吗？

我该如何判断鼻塞是否在缓解？

为什么区块链不直接把区块做得更大一些呢？

拥堵如何影响在……上进行开发的开发商Quicknode

延伸阅读

立即开始构建