Blockchain reliability league
Which chains actually stay up? Halts since 2024 from the public record, plus uptime we measure ourselves from each chain's own blocks.
Ranked by total time halted since January 2024; finality stalls count a quarter and degradations a tenth of their duration. Equal totals share a rank (=); within a tie, the longest run since the last halt comes first, and chains with only unconfirmed reports come last. "None on record" means we found no halt in postmortems or reliable reports; younger chains have simply had less time to fail. The first columns come from the historical record. The last column is measured by us from the chains' own blocks since 12 Nov 2026 (last 30 days once we have them).
| # | Chain | Since last halt | Halts since 2024 | Time halted | Other incidents | Uptime we measured |
|---|---|---|---|---|---|---|
| 1= | Dogecoin | None on record | 0 | — | — | 100% 3 days so far |
| 1= | Monad | None on record | 0 | — | — | 100% 3 days so far |
| 1= | Bitcoin | 4,998 days halt, Mar 2013 | 0 | — | — | 100% 3 days so far |
| 1= | TRON | 2,205 days halt, Nov 2020 | 0 | — | — | 100% 3 days so far |
| 1= | BNB Chain | 1,502 days halt, Oct 2022 | 0 | — | — | 100% 3 days so far |
| 1= | Aptos | 1,125 days halt, Oct 2023 | 0 | — | — | 100% 3 days so far |
| 1= | Arbitrum One | 1,067 days halt, Dec 2023 | 0 | — | — | 100% 3 days so far |
| 1= | ZKsync Era | 1,057 days halt, Dec 2023 | 0 | — | — | 100% 3 days so far |
| 1= | Scroll | None on record unconfirmed reports exist | 0 | — | — | 100% 3 days so far |
| 1= | Celestia | None on record unconfirmed reports exist | 0 | — | — | 100% 3 days so far |
| 1= | Hyperliquid | None on record unconfirmed reports exist | 0 | — | — | 100% 3 days so far |
| 12 | Polkadot | None on record | 0 | — | 1 degraded | 100% 3 days so far |
| 13 | Litecoin | None on record | 0 | — | 1 degraded | 100% 3 days so far |
| 14 | Zcash | None on record | 0 | — | 1 degraded | 100% 3 days so far |
| 15 | Ethereum | 1,284 days halt, May 2023 | 0 | — | 1 degraded | 100% 3 days so far |
| 16 | XRP Ledger | 650 days halt, Feb 2025 | 2 | 74 min longest 64 min | — | 100% 3 days so far |
| 17 | Cardano | None on record | 0 | — | 1 degraded | 100% 3 days so far |
| 18 | OP Mainnet | 895 days halt, Jun 2024 | 2 | 2.8 h longest 2.2 h | 2 degraded | 100% 3 days so far |
| 19 | Base | 142 days halt, Jun 2026 | 3 | 3.1 h longest 1.9 h | — | 100% 3 days so far |
| 20 | Osmosis | 2 days halt, Nov 2026 | 0 | — | — | 94.710% 3 days so far 1 halt, longest 4 h |
| 21 | Polygon PoS | 307 days finality stall, Jan 2026 | 0 | — | 3 finality stalls, 1 degraded | 100% 3 days so far |
| 22 | Solana | 1,013 days halt, Feb 2024 | 1 | 5.1 h longest 5.1 h | — | 100% 3 days so far |
| 23 | Avalanche C-Chain | 996 days halt, Feb 2024 | 1 | 5.4 h longest 5.4 h | — | 100% 3 days so far |
| 24 | TON | 532 days halt, Jun 2025 | 3 | 14 h longest 7.3 h | — | 100% 3 days so far |
| 25 | Starknet | 242 days halt, Mar 2026 | 3 | 16 h longest 11 h | 2 degraded | 100% 3 days so far |
| 26 | Linea | 38 days finality stall, Oct 2026 | 4 | 2.7 h longest 1.6 h | 1 finality stall | 100% 3 days so far |
| 27 | Sui | 169 days halt, May 2026 | 5 | 25 h longest 6.5 h | — | 100% 3 days so far |
| 28 | Cosmos Hub | 53 days halt, Sep 2026 | 2 | 29 h longest 25 h | — | 100% 3 days so far |
Monthly reliability report
Generated automatically from our measurements and the public record.
November 2026 to date
- 1 halt measured (since we started measuring on 12 Nov, to date): Osmosis. Total time halted 4 h; longest 4 h.
- Slowest hour: Osmosis on 13 Nov, 03:00 UTC, one block every 10.5 s on average against a typical 927 ms.
- Longest wait for a new block seen at a check (outside proof-of-work): Osmosis on 13 Nov, 4.3 h.
- Our sampling covered 80% of the minutes in that period.
- Public record: no halts or major disruptions reported in November 2026.
Copy as text
October 2026
- We were not measuring yet; figures below come from the public record only.
- Public record: Linea finality stall on 5 Oct 2026, 08:11 UTC (≈3 days).
Copy as text
Halts we detected
Measured by ChainReport since 12 Nov 2026. Start = timestamp of the last block before the stop; end = first new block (±1 minute).
| Chain | Type | Last block before halt | Duration | How we confirmed it |
|---|---|---|---|---|
| Osmosis | Halt | 12 Nov 2026, 22:57 UTC | ≈4 h | Confirmed by two independent sources: PublicNode and cosmos.directory |
Historical record, 2024 to today
Halts, finality stalls and serious degradations, from postmortems, status pages and news reports. Times are UTC; "≈" marks approximate times, and each row links its sources.
| Date | Chain | Type | Duration | What happened | Sources |
|---|---|---|---|---|---|
| 5 Oct 2026, 08:11 UTC | Linea | Finality stalled | ≈3 days | A prover/finalization issue delayed L1 finalization of Mainnet, while blocks kept being produced.Note on timesTimes are status-page post times. Proofs were generating normally from Oct 6 15:00 UTC, but finalization stayed delayed until the fix on Oct 8. Smaller L2->L1 finalization partial outages (Mar 23-26, May 23-24 and Jun 13-14, 2026) are not listed separately. | linea.statuspage.io |
| 22 Sep 2026, 11:00 UTC | Cosmos Hub | Halt | ≈25 h | Validators deliberately halted the Hub at height 33,086,740 after a malicious governance takeover on Neutron sent ~1.2M stolen ATOM to a Hub address; they restarted with a state change that recovered those funds.Note on timesAn intentional, coordinated halt, not a bug. The start time is approximate (~25 h before the 12:00 UTC Sep 23 restart). | cryptobriefing.com protos.com |
| 8 Aug 2026 | Linea | Halt | not reported | A full outage of Mainnet block production stopped network progression and finality until the sequencer was restarted.Note on timesThe official status page gives no start time. Block production was already restored by the 11:44 UTC 'monitoring' update, and the incident was closed Aug 10 03:56 UTC. Duration unknown. | linea.statuspage.io |
| 26 Jun 2026, 15:33 UTC | Base | Halt | ≈38 min | The same sequencer bug recurred with similar symptoms, a second unsafe-head stall in two days.Note on times15:33 UTC is the status alert and 16:11 UTC is when The Block reported resumption. The status page says production resumed at 15:47 UTC, and the postmortem summary gives ~20 min. | theblock.co |
| 25 Jun 2026, 16:03 UTC | Base | Halt | ≈1.9 h | The sequencer built an invalid block (#47,806,542) after a failed transaction left stale state in the block builder, which jammed all later block building (an 'unsafe head stall').Note on timesSequencing resumed 17:51 UTC; healthy by 17:58 UTC. The stale-journal cause is per secondary coverage of Base's postmortem (not opened). Base said it was unrelated to the Beryl hard fork. | news.bitcoin.com theblock.co |
| 3 Jun 2026, 05:28 UTC | Zcash | Degraded | ≈4 h | After the June 1 emergency Orchard-fix upgrade, block explorers and some services on faulty or unpatched nodes showed no new blocks for about four hours, freezing exchange deposits.Note on timesDisputed. Yellow and Cryptonews reported a real 4h halt (last block ~3,364,601/3,364,603 at ~05:27-05:28 UTC, next after ~09:30 UTC). ZODL and Helius's Mert Mumtaz said the chain stayed operational and blamed explorers on a faulty node. No official post-mortem. Classed as degraded, not halt. | kucoin.com yellow.com |
| 29 May 2026, 20:30 UTC | Sui | Halt | ≈5.8 h | When validators restarted for the fix, a latent bug that never persisted a failed DKG (randomness) status across restarts stalled the epoch change.Note on timesAbout 13:30-19:20 PT. Triggered by the restart for the v1.72 fix, not by address balances directly. | sui.io theblock.co |
| 29 May 2026, 12:00 UTC | Sui | Halt | ≈3.5 h | The same v1.72 underflow recurred because a different cancellation reason masked InsufficientFundsForWithdraw and bypassed the interim fix.Note on timesAbout 5:00-8:30 PT per the postmortem timeline. Elsewhere it says ~9:40 PT (16:40 UTC), so the duration may be up to ~280 min. | sui.io coindesk.com |
| 28 May 2026, 14:00 UTC | Sui | Halt | ≈6.5 h | The v1.72 address-balances feature exposed a gas-charging edge case: hybrid-gas transactions cancelled for insufficient funds still spent funds, creating a negative balance that crashed validators.Note on timesAbout 7:00-13:30 PT (PDT = UTC-7). Approximate. | sui.io theblock.co |
| 25 Apr 2026 | Litecoin | Degraded | ≈3 h | A zero-day in the MWEB layer let an invalid pegout be mined: upgraded nodes stalled while non-upgraded miners extended the invalid chain, causing a 13-block split/reorg (blocks 3,095,930-3,095,943) that F2Pool resolved.Note on timesNo clock times were published. The duration is approximate: 'more than three hours' per The Block, ~2h45m elsewhere. The primary postmortem (litecoin.com) returned 403. | theblock.co |
| 18 Mar 2026 | Starknet | Halt | ≈11 min | Mainnet was down for 11 minutes during planned maintenance for the 0.14.1 committer upgrade.Note on timesPlanned maintenance, not an unplanned failure. Only the 07:30 UTC entry time is known. Other planned upgrade windows (Jul 8, 2026, ~10 min; Oct 6, 2026, ~5-10 min) are excluded. | status.starknet.io |
| 4 Mar 2026, 23:04 UTC | OP Mainnet | Degraded | ≈10 min | Networking problems in the infrastructure linking block-building components caused slow block building and intermittent chain stalls.Note on timesMajor outage 23:04-23:07 UTC. | status.optimism.io |
| 14 Jan 2026, 14:22 UTC | Sui | Halt | ≈6.4 h | An edge case in consensus-commit logic under garbage-collection conditions made validators compute different commits, so more than 1/3 of stake signed a different checkpoint and certification stopped.Note on timesThe start is the last processed transaction (Suiscan via The Block); some reports say 14:52 UTC. The ~20:44 UTC end is from secondary reports in search results. The Sui postmortem says only 'about 6 hours'. | sui.io theblock.co |
| 12 Jan 2026, 16:00 UTC | Polygon PoS | Finality stalled | ≈2.7 h | A PrepareProposal/ProcessProposal mismatch in Heimdall made validators repeatedly reject valid proposals, stalling finalization until validators rolled back Heimdall.Note on timesThe postmortem gives a ~16:00 UTC start. The status page lists 17:40 and shows 'finality caught up' at 18:43 (timezone presumed UTC). | status.polygon.technology |
| 5 Jan 2026, 09:42 UTC | Starknet | Halt | ≈4.6 h | A blockifier bug restored state incorrectly when a nested call panicked and was caught, so sequencing was paused and the chain was reorged to drop the bad execution.Note on timesActivity from 09:24-09:42 UTC (18 min) was also reverted, so total impact was 09:24-14:17 (~293 min). The status page lists 09:53-14:02 with a revert to block 5187263. Press reported 2-4h. | starknet.io status.starknet.io |
| 17 Dec 2025, 18:00 UTC | Polygon PoS | Degraded | ≈14 h | Double-sealed blocks and span rotations with only two block producers online made lagging Bor nodes reject blocks, so many RPC nodes stalled while the chain itself stayed live.Note on times'Partial Bor Halt'. The hotfix was out by 22:00 UTC Dec 17 and most nodes were back in sync within hours; the end is the 'functionality restored' post. Block production continued. | status.polygon.technology |
| 4 Dec 2025, 02:50 UTC | Ethereum | Degraded | ≈4.4 h | After Fusaka, a Prysm v7 bug (PR 15965) made Prysm nodes regenerate old states when they processed attestations from out-of-sync peers, exhausting resources; participation fell to ~75%, but finality was never lost.Note on timesTimes are converted from the impacted epoch range 411439-411480 (42 epochs) given on Prysm's postmortem page (prysm.offchainlabs.com/docs/misc/mainnet-postmortems/; quoted in search results, returned 404 when opened). Blocks were missed (18.5% missed slots) but the chain kept finalizing. | cointelegraph.com cointelegraph.com |
| 21 Nov 2025, 08:00 UTC | Cardano | Degraded | ≈14 h | A malformed delegation transaction exploited a 2022 hash-deserialization bug: newer nodes accepted it and older nodes rejected it, splitting the chain in two until operators upgraded and the chains merged.Note on timesA chain split, not a halt: blocks continued on both forks. The 22:17 UTC reconvergence time is CryptoSlate's timeline as quoted in search results (page not opened, 403). Exchange pauses ran until about 02:10 UTC Nov 22. | intersectmbo.org decrypt.co |
| 28 Oct 2025, 10:38 UTC | Linea | Halt | ≈1.6 h | Mainnet block production stalled at block 24,992,514; no cause was disclosed.Note on timesTimes come from QuickNode's status page (first 'investigating' post to 'resolved'), so the real start may be earlier. No Linea post-mortem was found. | status.quicknode.com |
| 9 Oct 2025 | Starknet | Degraded | not reported | Aerospike database hotspots from preconfirmed-block fetches stalled the committer, and a transaction execution-time misconfiguration caused repeated retry loops, producing a series of short stalls.Note on timesSeries of downtimes, mostly a few minutes each, with three lasting 10-20 minutes. The official forum post gives no per-event timestamps. | community.starknet.io |
| 10 Sep 2025, 07:11 UTC | Polygon PoS | Finality stalled | ≈10 h | A bug in Heimdall's milestone (finality-gadget) process and Bor/Erigon left some nodes unable to progress, delaying finality 10-15+ minutes; a Heimdall hard fork at ~15:00 UTC (Bor v2.2.11-beta2, Heimdall v0.3.1) restored it.Note on timesBlocks continued. The window runs from the first status post (07:11) to 'consensus finalization restored' (17:28, times presumed UTC). Finality was delayed rather than fully stopped for the whole window. Marked resolved Sep 11 07:00. | status.polygon.technology |
| 10 Sep 2025, 05:25 UTC | Linea | Halt | ≈67 min | The mainnet sequencer stopped producing blocks after block 23,145,386 on the day of the LINEA token launch; Linea said it found and fixed the issue but did not disclose a root cause.Note on timesTimes converted from 1:25-2:32 a.m. ET (EDT). Other coverage cites a 46-minute gap between blocks 23,145,386 and 23,145,387. An aggregator also lists a ~11h incident on Sep 11 at block 23,202,037, which could not be verified. | theblock.co |
| 2 Sep 2025, 02:20 UTC | Starknet | Halt | ≈11 h | After the v0.14.0 (Grinta) upgrade, Ethereum node failures gave the three sequencers different views of L1, a manual intervention produced conflicting blocks, and a blockifier bug in reprocessing reverted L1-to-L2 messages forced two reorgs.Note on timesOfficial report: three halts within the window, 02:20-03:57, 04:37-07:42 and 10:43-13:41 UTC (~460 min actually halted, ~9h degraded overall). Reorg 1 rolled back to block 1,960,612 and reorg 2 rolled back from 1,962,681, reverting ~1.5h of transactions. Early Cointelegraph coverage said 2h44m (02:23-04:36), which was only the first phase. | starknet.io cointelegraph.com |
| 5 Aug 2025, 06:07 UTC | Base | Halt | ≈33 min | The active sequencer lagged under load, and op-conductor handed leadership to a standby that was still being provisioned and could not build blocks; engineers paused Conductor and elected a healthy sequencer.Note on times06:07-06:40 UTC is per most secondary sources and status trackers. The Block says 'around 1:07 a.m.' with no timezone. | theblock.co cointelegraph.com |
| 30 Jul 2025, 14:00 UTC | Polygon PoS | Finality stalled | ≈61 min | A validator exit triggered a Heimdall bug that halted consensus/chain progression for about an hour; Bor kept producing blocks but RPC providers had to resync.Note on timesSources conflict. The Block (citing Polygon) says Bor blocks continued, detected shortly before 10:00 ET and resolved 11:01 ET. Blocmates says new blocks halted ~50 min from about 12:07 UTC. | theblock.co blocmates.com |
| 1 Jun 2025, 12:51 UTC | TON | Halt | ≈40 min | An error in processing the masterchain dispatch queue stopped block production; a quick patch to a few masterchain validators restored it.Note on times12:51 UTC is when TON announced the issue, not necessarily when the halt began; the ~40 min is from identification to restoration. | cointelegraph.com |
| 4 Feb 2025 | XRP Ledger | Halt | ≈64 min | Consensus kept running but validations were not published, so the network drifted apart and stopped validating ledgers at height 93,927,173 until it re-converged.Note on timesClock times are uncertain. The Block says 'late Tuesday' with no times. A secondary report says the ledger restarted at 10:58 UTC on Feb 4. The duration is about 64 min. | theblock.co |
| 25 Nov 2024, 13:39 UTC | XRP Ledger | Halt | 10 min | A caching-layer bug in rippled could return an inconsistent result type and crash servers; many nodes crashed and restarted, so no transactions were processed for about 10 minutes (fixed in rippled 2.3.0). | dev.to |
| 21 Nov 2024, 09:15 UTC | Sui | Halt | ≈2.5 h | An assert in congestion-control (transaction scheduling) code crashed every validator when the estimated execution cost was zero, leaving them in a crash loop until v1.37.4 shipped.Note on timesThe postmortem gives about 1:15-3:45 am PT (PST = UTC-8). Times are approximate. | sui.io theblock.co |
| 17 Sep 2024, 07:58 UTC | Polkadot | Degraded | ≈14 min | The runtime 1.3.0 upgrade migrated HostConfiguration at the end of the block, so a runtime API read corrupted values and the candidate-backing subsystem stalled; relay finality lagged about 70 blocks and parachains nearly stopped.Note on timesRelay finality caught up around 08:12 UTC; parachain collators needed manual restarts from about 08:30 UTC. The relay chain did not fully halt. | forum.polkadot.network |
| 28 Aug 2024, 19:19 UTC | TON | Halt | 6.2 h | A second halt from heavy load attributed to DOGS minting; validators had to update and restart their nodes. | theblock.co |
| 27 Aug 2024, 22:11 UTC | TON | Halt | ≈7.3 h | Load from the DOGS memecoin left several validators unable to clean old transactions from their databases, so they lost consensus; validators restarted to recover.Note on timesStart per Tonscan via The Block. The 05:30 UTC end is TON Foundation's announcement as quoted in search results, and the coordinated restart was at 04:00 UTC. Reported durations range from 3 to 7 h. | theblock.co |
| 5 Jun 2024, 19:00 UTC | Cosmos Hub | Halt | ≈4 h | The Cosmos Hub v17 upgrade, which also patched a Liquid Staking Module vulnerability, stopped block production for about four hours.Note on timesApproximate: the source says around 22:00 Kyiv/MSK time (UTC+3), so about 19:00 UTC, with a ~4 h halt. The exact end time is not stated. | forklog.com |
| 3 Jun 2024, 20:00 UTC | OP Mainnet | Halt | ≈32 min | Unsafe-head stall on OP Mainnet (sequencer stopped advancing); no root cause published.Note on timesStatus page times; the fix was applied at 20:04 UTC. | status.optimism.io |
| 2 Jun 2024 | Linea | Halt | not reported | Linea deliberately paused its sequencer between blocks 5081800 and 5081801 to stop the Velocore DEX exploiter from bridging funds out.Note on timesIntentional halt, not a failure. Sources give no UTC start or resume time. The block explorers (lineascan) refused automated fetches, so the block timestamps could not be checked. Date approximate (exploit Jun 2, 2024; coverage Jun 3). | theblock.co |
| 26 May 2024, 21:36 UTC | OP Mainnet | Degraded | ≈82 min | Instability in L1 nodes made blob data periodically unavailable to the sequencer, so transaction sequencing degraded.Note on timesPartial outage 21:44-22:16 UTC; not a full halt. | status.optimism.io |
| 4 Apr 2024 | Starknet | Degraded | ≈4 h | A rounding-error bug forced a block reorganization, leaving a roughly four-hour gap in blocks and briefly rejecting or reverting some transactions.Note on timesApproximate. Starkscan shows a ~4h gap between blocks 630028 and 630029 on Apr 4 (the article's lead says Apr 5). Starknet's status page showed no outage, and Starknet said block production continued, so whether this was a true halt is disputed. No UTC times given. | cointelegraph.com |
| 23 Feb 2024, 11:13 UTC | Avalanche C-Chain | Halt | ≈5.4 h | A gossip/mempool-management bug, triggered by an inscription wave, filled validators' bandwidth with useless gossip so consensus pull queries stalled; fixed in AvalancheGo v1.11.1.Note on timesStart from Forklog (11:13 UTC); finalization resumed 11:36 ET per The Block. The Block calls it ~6 hours; others say 2-4 hours. | theblock.co forklog.com |
| 15 Feb 2024, 05:04 UTC | OP Mainnet | Halt | ≈2.2 h | Unsafe-head stall on OP Mainnet (sequencing stopped); no root cause published on the status page.Note on timesStatus-page window; unrelated to the planned withdrawal pause the next day. | status.optimism.io |
| 6 Feb 2024, 09:53 UTC | Solana | Halt | ≈5.1 h | A bug in the validator's JIT-compilation cache (Agave v1.17) left nearly all validators stuck in a recompile loop, so voting and consensus stopped until a patched v1.17.20 restart.Note on timesTimes from the Anza postmortem as quoted in search results (postmortem page returned 429 when opened); Cointelegraph (opened) gives 10:22 UTC as the start. | solana.com cointelegraph.com |
Before 2024
The most recent halt we could find for chains with none since 2024.
- Bitcoin 11 Mar 2013 · Chain fork: v0.8 (LevelDB) nodes accepted a large block that pre-0.8 (BerkeleyDB lock limit) nodes rejected; pools downgraded to 0.7 to reorg it away. Before that, the Aug 2010 value-overflow incident. No halt in 2024-2026. Source
- Ethereum 12 May 2023 · Last actual loss of finality: 3 epochs on May 11 and 9 epochs (~1 h) on May 12, 2023, from Prysm/Teku resource exhaustion on old attestations. Block production has never halted since the Merge. Dec 2025 Prysm/Fusaka was a near-miss (75% participation), not a finality loss. Source
- Arbitrum One 15 Dec 2023 · An Arbitrum One sequencer and feed stall during an inscription surge, a major outage of ~4 h (16:57-21:04 UTC per the status page). No Arbitrum One halt found for 2024-2026 (a Mar 23 2026 'block production halted' incident was Arbitrum Sepolia, a testnet). Source
- ZKsync Era 25 Dec 2023 · zkSync Era block production stopped for about 5 h after a server bug tripped an automated safety mechanism (there was also a halt on Dec 16, 2023). No 2024-2026 halt found. Source
- Scroll Unconfirmed · No confirmed Scroll mainnet halt. Third-party RPC monitors (QuickNode and Alchemy, via isdown.app) reportedly logged stalls on 2025-10-28 (block 23,915,173) and 2026-01-14 (block 28,074,287, ~40 min), but those pages returned 403 and Scroll's status page is deleted, so these are unverified. Source
- BNB Chain 6 Oct 2022 · BNB Smart Chain validators halted the chain for ~9 h after the BSC Token Hub bridge exploit (~2M BNB forged) and restarted it at 2:53 a.m. ET on Oct 7. No 2024-2026 halt found (a June 4 2026 'BNB chain stalled' notice was a Dune indexing outage). Source
- Aptos 18 Oct 2023 · Block production stopped for more than 5 h (from ~23:11 UTC, block 104,621,314) during the first-anniversary Graffio canvas event; no root cause was disclosed. No 2024-2026 halt found. Source
- TRON 2 Nov 2020 · Super representatives halted block production for ~2-3.5 h after a 'malicious contract' during the v4.1 upgrade. No later mainnet halt found. Source
- Osmosis 8 Jun 2022 · Validators halted the chain at block 4,713,064 after a v9 Nitrogen liquidity-pool bug let users withdraw ~50% more than they deposited (~$5M lost). No 2024-2026 halt found. Source
- Polkadot None known (relay-chain halt) · No Polkadot relay-chain block-production halt found. In range there is the Sep 17 2024 finality lag (degraded, in incidents). Kusama (not Polkadot) had finality/block stalls in 2024-2025, and Polkadot parachains, People Chain and Asset Hub had stalls in 2026 with the relay chain unaffected. Source
- Dogecoin None known · No Dogecoin network halt found. The Dec 2024 'DogeReaper' bug crashed ~69% of public nodes, but blocks kept being mined. (Dogechain is a separate project.) Source
- Celestia Unconfirmed · No confirmed Celestia mainnet halt. A Fireblocks status entry (via isdown.app, not opened) logged a 'Celestia halt' of ~11 h and slow block production on 2025-11-24, the same day as the v6 'Matcha' mainnet upgrade (12:33 UTC, height 8,662,012). Not verified as a chain halt. Source
- Hyperliquid Unconfirmed · No confirmed HyperCore/HyperEVM chain halt. QuickNode (via isdown.app, 403 when opened) reportedly logged a HyperEVM network-wide stall at block 8,816,123 on 2025-07-19 (~2 h, tied to a scheduled network upgrade). The Jul 29 2025 outage was API/order execution only (status showed L1 fine), and the Apr 2 2026 PeckShield 'HyperEVM failure' appears to have been explorer data lag. Source
- Monad None known · No Monad mainnet halt found. Only regional RPC degradation (QuickNode/Chainstack LAX, Jun 2026) and a Mar 2025 testnet stall. Mainnet launch date not verified here. Source
- Starknet 15 Nov 2023 · Block production stalled for about 50 minutes (a 40-minute gap, then a 10-minute gap) from 10:00 GMT, when a transaction influx overloaded the gateway's single deduplicator. Source
- Linea None known · Linea mainnet launched in July 2023; no halt found before 2024. Source
- Zcash None known · No pre-2024 chain halt known. The most recent serious incident was the 2022 'sandblasting' spam of low-fee, high-output shielded transactions, which bloated the chain and crippled wallet sync. It led to the ZIP-317 fee change (ZIP created 2022-08-15, live with zcashd 5.5.0 in April 2023). Source
How this works and where the data comes from
What we measure. Every minute a server job asks each chain for its newest block and reads the timestamp the chain itself put in that block. "Last block" is now minus that timestamp, so it is accurate even if our own checks run late. "Last hour" is the average spacing between blocks over the past hour, also from block timestamps.
When we call a halt. A chain counts as halted when no new block has appeared for much longer than normal. The limit is set per chain: 2 minutes for most fast chains, 5–7 minutes for Scroll, ZKsync and Cardano, 14 minutes for Dogecoin, 30 minutes for Litecoin and 2 hours for Bitcoin, where long gaps happen by chance. Before declaring a halt we ask a second, independent data provider. If one provider is stale and the other is fresh, the chain is fine and we log the provider as lagging. For Solana, Sui and Cardano we have only one free source, so we wait for three stale checks in a row. Start time is the last block before the stop; end time is the first new block, within about a minute.
Finality. For Ethereum we read the last finalized checkpoint from the beacon chain (normally 2 epochs, about 13–19 minutes behind). For other EVM chains and Polkadot we read the newest "finalized" block from the node. On rollups (Base, Arbitrum, OP, ZKsync, Scroll), "finalized" means settled on Ethereum, which takes longer by design: hours on ZKsync. Free Solana endpoints do not report commitment levels reliably, so we do not show Solana finality.
Limits. We have been measuring since 12 Nov 2026; uptime figures cover only that period and say so. If our job misses minutes, a halt that started and ended entirely inside that gap would show up only as a slow hour in the chart. The historical record before our measurements comes from postmortems and news reports, with approximate times marked "≈". Hyperliquid is measured through HyperEVM, which stops whenever the HyperCore L1 stops.
Sources: mempool.space; Blockstream Esplora; PublicNode; dRPC; PublicNode Ethereum beacon API; Base public RPC; Arbitrum public RPC; Optimism public RPC; ZKsync public RPC; Scroll public RPC; Linea public RPC; BNB Chain public RPC; Avalanche public API; Sui GraphQL (Mysten Labs); Aptos Labs API; TronGrid; cosmos.directory; Parity public RPC; Koios; TON Center; TonAPI; XRPL Cluster; Ripple public server; litecoinspace.org; BlockCypher; Zelcore Blockbook; Hyperliquid public RPC; Monad public RPC; Monad Foundation RPC; Historical record: project postmortems, status pages, The Block, CoinDesk, Cointelegraph, Decrypt (linked per row)
Information only, not financial advice. Data comes from public sources and can be delayed or wrong; check the primary source before acting.