<oembed><type>rich</type><version>1.0</version><author_name>npub1s4lj77xuzcu7wy04afcr487f0r3za0f8n2775xrpkld2sv639mjqsd44kw</author_name><author_url>https://nostr.ae/npub1s4lj77xuzcu7wy04afcr487f0r3za0f8n2775xrpkld2sv639mjqsd44kw</author_url><provider_name>njump</provider_name><provider_url>https://nostr.ae</provider_url><html>📅 Original date posted:2015-05-31&#xA;📝 Original message:On Sat, May 30, 2015 at 9:31 PM, Chun Wang &lt;1240902 at gmail.com&gt; wrote:&#xA;&gt;&#xA;&gt; If someone propagate a 20MB block, it will take at best 6 seconds for&#xA;&gt; us to receive to verify it at current configuration, result of one&#xA;&gt; percent orphan rate increase.&#xA;&#xA;&#xA;That orphan rate increase will go to whoever is producing the 20MB blocks,&#xA;NOT you.&#xA;&#xA;&#xA;Or, we can mine the next block only on&#xA;&gt; the previous block&#39;s header, in this case, the network would see many&#xA;&gt; more transaction-less blocks.&#xA;&gt;&#xA;&#xA;Are you sure that is the best strategy? If a big block is slow to&#xA;propagate, I suspect it will be better to punish the miner that created it&#xA;by refusing to build on it until it has been fully validated.&#xA;&#xA;I&#39;ll try to find time to run a couple of simulations.&#xA;&#xA;&#xA;&#xA;&gt;&#xA;&gt; Our orphan rate is about 0.5% over the past few months. If the network&#xA;&gt; floods 20MB blocks, it can be well above 2%. Besides bandwidth, A 20MB&#xA;&gt; block could contain an average of 50000 transactions, hundred of&#xA;&gt; thousands of sigops, Do you have an estimate how long it takes on the&#xA;&gt; submitblock rpccall?&#xA;&gt;&#xA;&#xA;I can benchmark it. It should be pretty fast, and sipa has a couple of&#xA;patches pending to make the UTXO cache much faster.&#xA;&#xA;It can be fast because the vast majority of the work of validating all&#xA;those transactions can happen as they are received into the memory pool.&#xA;&#xA;&#xA;&gt; For references, our 30Mbps bandwidth in Beijing costs us 1350 dollars&#xA;&gt; per month.&#xA;&#xA;&#xA;You should be able to handle 20MB blocks no problem; if I round up to 100MB&#xA;per block that works out to 1.3Mbps.&#xA;&#xA;We also use Aliyun and Linode cloud services for block&#xA;&gt; propagation. As of May 2015, the price is 0.13 U.S. dollars per GB for&#xA;&gt; 100Mbps connectivity at Aliyun.&#xA;&#xA;&#xA;That speed will handle 20MB blocks no problem.&#xA;&#xA;If each 20MB block is 100MB of data up/down the wire (I&#39;m vastly&#xA;over-estimating, after optimization it should be 40MB) then you&#39;ll be&#xA;paying...uhhh:&#xA;&#xA;0.1 GB / block-data-on-wire * 144 blocks/day * 30.5 days/month * 0.13 $ /&#xA;GB = $57&#xA;&#xA;Less than $2 per day in bandwidth, surely you can afford that.&#xA;&#xA;&#xA;&gt; For a single cross-border TCP&#xA;&gt; connection, it would be certainly far slower than 12.5 MB/s.&#xA;&#xA;&#xA;That&#39;s OK, you&#39;ll 1.3Mbps or less.&#xA;&#xA;&#xA;&gt; I think we can accept 5MB block at most.&#xA;&gt;&#xA;&#xA;Are you worried about paying too much, or do 20MB blocks &#34;feel like too&#xA;much&#34; ?&#xA;&#xA;-- &#xA;--&#xA;Gavin Andresen&#xA;-------------- next part --------------&#xA;An HTML attachment was scrubbed...&#xA;URL: &lt;http://lists.linuxfoundation.org/pipermail/bitcoin-dev/attachments/20150531/41d199dc/attachment.html&gt;</html></oembed>