Making file systems with a billion files is interesting for feeling out scaling issues.<\/p>\n\n\n\n
The Intro post for file systems with a billion files, with a table of contents<\/a>. This is yet another way to make file systems with a billion files<\/a>. <\/p>\n\n\n\n While working on the upcoming archiving and compression post, with various obstacles, yet another method for making those file systems came to mind: running multiple multitouch methods in parallel. Spoilers: It\u2019s the fastest method for making file systems with a billion files that I\u2019ve run.<\/p>\n\n\n\n\n\n\n\n I worked from a Raspberry Pi 4, with 4 GB RAM, running Debian 12 (bookworm).\nIt has four cores.\nFor storage – this is new – I connected the Pi to my NAS.<\/p>\n\n\n\n While gathering data for the upcoming archiving and compression post, the first hard drive, a 2017-era Seagate USB hard drive, flaked out.\nWhen connected to my Pi, the drive would go offline, with kernel log messages involving \u201cover-current change\u201d – which might point at the power supply for the Pi.\nThis hard drive\u2019s shelved for now.<\/p>\n\n\n\n The second spare drive I reached for was a 2023-era WD USB hard drive.\nIt was also an SMR drive, and the performance was surprisingly awful.\nAs in, I went looking for malware.\nPerhaps it just needs some time for internal remapping and housekeeping, but it\u2019s off to the side for now.<\/p>\n\n\n\n I connected the Pi to my NAS, with redundant CMR hard drives, over Gigabit Ethernet.<\/p>\n\n\n\n Before, I\u2019d placed a few scripts in a fork for create-empty-files; I\u2019m adding more and it feels like tangential clutter. So, I started a I realized I could run multiple multitouch in parallel.<\/p>\n\n\n\n There’s a rule of thumb for parallel processes: the number of processor cores, plus one. I benchmarked with a million files, and one to ten processes; the peak was five, matching that expectation.<\/p>\n\n\n\n It populated a billion files in under seven hours, a new personal record!<\/p>\n\n\n\n Here are how the fastest generators ran, against NAS storage:<\/p>\n\n\n\n Against the first USB drive, Rust & ext2 was slower than ext4.\nThey switched places here, which surprised me.<\/p>\n\n\n\n Probably archiving and compression – it’s already much longer, with graphs.<\/p>\n","protected":false},"excerpt":{"rendered":" about Making file systems with a billion files is interesting for feeling out scaling issues. The Intro post for file systems with a billion files, with a table of contents. This is yet another way to make file systems with a billion files. While working on the upcoming archiving and compression post, with various obstacles, […]<\/p>\n","protected":false},"author":1,"featured_media":0,"comment_status":"open","ping_status":"open","sticky":false,"template":"","format":"standard","meta":{"footnotes":""},"categories":[1],"tags":[17],"class_list":["post-213","post","type-post","status-publish","format-standard","hentry","category-uncategorized","tag-billion-files"],"_links":{"self":[{"href":"https:\/\/pronoiac.org\/misc\/wp-json\/wp\/v2\/posts\/213"}],"collection":[{"href":"https:\/\/pronoiac.org\/misc\/wp-json\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/pronoiac.org\/misc\/wp-json\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/pronoiac.org\/misc\/wp-json\/wp\/v2\/users\/1"}],"replies":[{"embeddable":true,"href":"https:\/\/pronoiac.org\/misc\/wp-json\/wp\/v2\/comments?post=213"}],"version-history":[{"count":6,"href":"https:\/\/pronoiac.org\/misc\/wp-json\/wp\/v2\/posts\/213\/revisions"}],"predecessor-version":[{"id":219,"href":"https:\/\/pronoiac.org\/misc\/wp-json\/wp\/v2\/posts\/213\/revisions\/219"}],"wp:attachment":[{"href":"https:\/\/pronoiac.org\/misc\/wp-json\/wp\/v2\/media?parent=213"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/pronoiac.org\/misc\/wp-json\/wp\/v2\/categories?post=213"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/pronoiac.org\/misc\/wp-json\/wp\/v2\/tags?post=213"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}hardware<\/h2>\n\n\n\n
environmental changes<\/h3>\n\n\n\n
new repo<\/h2>\n\n\n\n
billion-file-fs<\/code> repo<\/a>. I plan to put everything in this post there; see the scripts<\/code> directory<\/a>.<\/p>\n\n\n\nparallel multitouch<\/h2>\n\n\n\n
processes<\/th> time<\/th> files \/ sec<\/th><\/tr><\/thead> 1<\/td> 1m57s<\/td> 8490<\/td><\/tr> 2<\/td> 1m5s<\/td> 15200<\/td><\/tr> 3<\/td> 50s<\/td> 19700<\/td><\/tr> 4<\/td> 45.2s<\/td> 22000<\/td><\/tr> 5<\/td> 43.7s<\/td> 22800<\/td><\/tr> 6<\/td> 44.1s<\/td> 22600<\/td><\/tr> 7<\/td> 45.0s<\/td> 22100<\/td><\/tr> 8<\/td> 44.8s<\/td> 22200<\/td><\/tr> 9<\/td> 44.8s<\/td> 22200<\/td><\/tr> 10<\/td> 45.0s<\/td> 22100<\/td><\/tr><\/tbody><\/table><\/figure>\n\n\n\n how did it do?<\/h3>\n\n\n\n
method<\/th> time<\/th> space used<\/th> files\/sec<\/th><\/tr><\/thead> multitouch on ext4<\/td> 18hr36min<\/td> 260GB<\/td> 14900<\/td><\/tr> Rust program on ext4<\/td> 14hr14min<\/td> 278GB<\/td> 19500<\/td><\/tr> Rust program on ext2<\/td> 12hr48min<\/td> 276GB<\/td> 21700<\/td><\/tr> parallel multitouch on ext4<\/td> 6hr41min<\/td> 260GB<\/td> 41500<\/td><\/tr><\/tbody><\/table><\/figure>\n\n\n\n what\u2019s next?<\/h2>\n\n\n\n