What's causing the rapid growth of UTXO set/chainset since April 2023?

8 replies 168 views
darkninjaMember
Posts: 8 · Reputation: 100
#1Apr 25, 2020, 03:57 AM
Since April 2023, the size of the chainset or UTXO set has been skyrocketing compared to previous years. What’s up with that? To give you an idea, the size can differ across different node setups, but checking one node shows some crazy numbers. We had a peak of 34% growth from 2019-2020, and on average, it was around 15% per year from 2019 to 2023. But this last year? It jumped by a whopping 100%, especially ramping up since roughly the last week of April. Interestingly, the blockchain size hasn't really followed suit: Even though the blockchain size did see some changes since February 2023 (maybe because of Ordinals?), with the average block size apparently going from about 1.2MB to 1.7MB, it’s still nothing compared to how fast the UTXO set is growing. Plus, the start dates don’t quite match up.
3 Reply Quote Share
atlas_2015Senior Member
Posts: 151 · Reputation: 999
#2Apr 25, 2020, 10:00 AM
This is due to inscriptions like BRC-20.  Notice how many (worthless) outputs are created as children of this transaction: https://mempool.space/tx/b97c394bb385792f002fd52238fcdfc2513582a45b67ff4639c0c5e4aa844bb2
4 Reply Quote Share
im_apeHero Member
Posts: 629 · Reputation: 3824
#3Apr 25, 2020, 12:58 PM
I warned about the severity of the attack under the codename Ordinals increasing more than 7 months ago: Considering the little to no effort that is being put into fixing the exploit these attackers are using, the situation has not improved so the rate of the spam is still high and unchanged as it can be seen from the slope in the updated chart below: https://statoshi.info/d/000000009/unspent-transaction-output-set?orgId=1&from=1682080884000&to=1704344007000
2 Reply Quote Share
hash_bossLegendary
Posts: 1166 · Reputation: 5261
#4Apr 25, 2020, 02:01 PM
Yes, it's caused by Ordinals (mainly BRC-20 which depends on Ordinals). I've observed this problem back on June 2023. This trend still on-going and already slow down IBD process for device with 8GB RAM (such as Raspberry Pi 4 8GB) or less.
0 Reply Quote Share
atlas_2015Senior Member
Posts: 151 · Reputation: 999
#5Apr 25, 2020, 07:37 PM
There's no "exploit" and it's not an "attack".  These transactions are valid and pay fees to miners which secure the network.
4 Reply Quote Share
darkninjaMember
Posts: 8 · Reputation: 100
#6Apr 26, 2020, 01:21 AM
Thanks. Is there someplace to see per-block stats, like how many transactions/outputs are Inscriptions, or the percentage of block data? Did BRC-20 and "general" Inscriptions both become possible from day one, once the feature was introduced? FWIW, these seem to be the number of Inscriptions per month: 2022/12 3 2023/01 445 2023/02 218K 2023/03 444K 2023/04 2.03M (the pace picked up more on 4/22) 2023/05 7.66M 2023/06 3.99M 2023/07 6.45M 2023/08 6.83M 2023/09 7.28M 2023/10 1.81M 2023/11 8.34M 2023/12 7.64M BTW, these things are ridiculous.
0 Reply Quote Share
darkninjaMember
Posts: 8 · Reputation: 100
#7Apr 26, 2020, 07:18 AM
Curiously, the UTXO set size has stabilized since about 5/2024, with a minor peak on 12/2024. The rate of new inscriptions has slowed down, but I'm not sure that's the only reason. There's a local increase in the rate 2 months ago but it hasn't negatively affected the UTXO set size, which actually dropped a bit. Inscriptions per month: 2024/01 6.51M 2024/02 3.62M 2024/03 4.08M 2024/04 3.67M 2024/05 0.61M 2024/06 0.93M 2024/07 1.52M 2024/08 1.46M 2024/09 1.14M 2024/10 0.68M 2024/11 1.73M 2024/12 2.51M 2025/01 2.84M 2025/02 4.19M 2025/03 3.76M 2025/04 ... https://statoshi.info/d/000000009/unspent-transaction-output-set?orgId=1&from=1672524000000&to=1745009999000
4 Reply Quote Share
im_lynxHero Member
Posts: 515 · Reputation: 2161
#8Apr 27, 2020, 06:37 AM
It may not contribute much to the UTXO set, but I don't know how much. For fun I've a watch-only wallet in Bitcoin Core assembled with combo(public_key) descriptors for blocks that are presumably attributed to Satoshi. Below I'm only looking at the Genesis block from which we know that Satoshi definitely has "mined" it. I exported with  listsinceblock  all UTXOs of the Genesis block, including the unspendable 50BTC of the coinbase transaction. The last exported UTXOs from my sample set were confirmed today at blocktime 2025-04-20T05:18:45Z UTC. Below table lists the numbers of UTXOs thrown for whatever reasons at the Genesis block. I don't expect any of them to ever move again (see below). Year | UTXOs2009 | 12010 | 02011 | 192012 | 552013 | 7932014 | 822015 | 552016 | 682017 | 762018 | 3342019 | 4242020 | 7922021 | 4812022 | 4442023 | 30022024 | 414382025 | 5706Total | 53770 There was clearly some madness going on in 2024 and I would expect on average that 2025 may end with approx. 18k new UTXOs in total. Many UTXOs are just 546sat dust or roughly multiples of it, some are 10000sat, rarely we see "donation like" transactions of a few hundred thousands sats and very rarely huge show-off "donations" or whatever such large transfers should resemble. Sigh... It's unlikely that Satoshi will come for the rescue to consolidate all that UTXO set bloat. Though it would be fun, for sure.
0 Reply Quote Share
im_apeHero Member
Posts: 629 · Reputation: 3824
#9Apr 28, 2020, 09:02 PM
Try checking if there is any link between these outputs and the Ordinals Fever, it may explain the reason for this kind of increase of spam transactions in 2024.
1 Reply Quote Share

Related topics