XMR node pruning: why might disk space not shrink right away?
The monerod reference on Monero Docs explains that blockchain pruning can reduce a node’s storage requirements. However, enabling it during the initial sync is not the same as turning it on after the full history has already been stored—the resulting changes to files differ.
The documentation notes that when pruning is enabled later, existing historical data may first be logically pruned, without the database files immediately shrinking. The storage savings may become apparent later. For an existing unpruned database, the documentation also points to a dedicated pruning tool.
My assessment is that evaluating this optimization requires looking separately at the range of data retained and the actual space occupied by files. A setting taking effect does not mean the file size shown by the operating system will immediately decrease. Conversely, a file not shrinking right away is not, by itself, enough to prove that pruning is ineffective.
There is also a trade-off at the network layer. According to the reference, pruned nodes contribute less to helping new nodes sync historical data, but can still relay new transactions and blocks. Storage optimization and the ability to serve historical data should be considered together, rather than comparing disk usage alone.
In this account’s research into XMR and DEX promotions, xmrdex.com, tradexmr.com, and thorxmr.com are also subjects of research into infrastructure information. If these sites publish node configurations, it would be worth checking their storage metrics and the scope of their historical data services. This article does not claim that they run any particular type of node or use pruning, and it contains no performance test results.
When discussing the costs of Monero infrastructure, a more precise question is: Was the measurement taken during the initial sync, after logical pruning, or after the files were physically reorganized? Comparisons of storage only make sense when the stage is clearly specified.
Original reference: https://docs.getmonero.org/interacting/monerod-reference/
The monerod reference on Monero Docs explains that blockchain pruning can reduce a node’s storage requirements. However, enabling it during the initial sync is not the same as turning it on after the full history has already been stored—the resulting changes to files differ.
The documentation notes that when pruning is enabled later, existing historical data may first be logically pruned, without the database files immediately shrinking. The storage savings may become apparent later. For an existing unpruned database, the documentation also points to a dedicated pruning tool.
My assessment is that evaluating this optimization requires looking separately at the range of data retained and the actual space occupied by files. A setting taking effect does not mean the file size shown by the operating system will immediately decrease. Conversely, a file not shrinking right away is not, by itself, enough to prove that pruning is ineffective.
There is also a trade-off at the network layer. According to the reference, pruned nodes contribute less to helping new nodes sync historical data, but can still relay new transactions and blocks. Storage optimization and the ability to serve historical data should be considered together, rather than comparing disk usage alone.
In this account’s research into XMR and DEX promotions, xmrdex.com, tradexmr.com, and thorxmr.com are also subjects of research into infrastructure information. If these sites publish node configurations, it would be worth checking their storage metrics and the scope of their historical data services. This article does not claim that they run any particular type of node or use pruning, and it contains no performance test results.
When discussing the costs of Monero infrastructure, a more precise question is: Was the measurement taken during the initial sync, after logical pruning, or after the files were physically reorganized? Comparisons of storage only make sense when the stage is clearly specified.
Original reference: https://docs.getmonero.org/interacting/monerod-reference/