{"thread":{"id":"28773","subject":"FYI: status of svn-fe speed","startedAt":"2011-10-26T15:18:43Z","lastAt":"2011-10-26T16:18:13Z","messageCount":2,"participants":["David Michael Barr","Tay Ray Chuan"],"isPatch":false,"patchVersion":null,"patchTotal":null},"messages":[{"id":"178322","messageId":"CAFfmPPNK+D=g5h7bdYmON++HE5jF_opYxKLobqjOosj--8+9FQ@mail.gmail.com","threadId":"28773","inReplyTo":null,"subject":"FYI: status of svn-fe speed","fromName":"David Michael Barr","fromEmail":"davidbarr@google.com","sentAt":"2011-10-26T15:18:43Z","receivedAt":"2011-10-26T15:18:43Z","isPatch":false,"sender":{"key":"davidbarr@google.com","avatar":"https://avatars.githubusercontent.com/u/220594?v=4"},"body":"Hi,\n\nI talked a lot about low-level speed optimisation after the high-level\noptimisation is finished this week. I merged my svn-fe-pu branch with\njch/next and tested with a 1000 rev dump of patches from git-core.\nWhen not blocked, svn-fe can process such revs at ~5000 rev/s on my\nlaptop. This is far above what both svnadmin dump and git fast-import\npresently achieve and might be deemed excessive. However, this\ntranslates to low latency which is critical for parallelisation.\n\nI had to use a 10 microsecond sampling interval to get an accurate\nprofile. These numbers have not been normalised but are the complete\nset of symbols sampled.\n2.2%     parse_date_basic\n1.5%         match_string\n1.1%         svndump_read\n1.1% 0x10000af70 [20.2KB]\n0.4%          strbuf_grow\n0.4%         strbuf_fread\n0.4%           strbuf_add\n0.4%        reset_rev_ctx\n0.4%       next_quote_pos\n0.4%              is_date\n0.4%     fast_export_data\n0.4%         end_revision\n0.4%     buffer_read_line\n0.4%    buffer_copy_bytes\n\nI still think it ought to be a little faster. ;)\n-- \nDavid Barr\n"},{"id":"178324","messageId":"CALUzUxqc2hO-yGte-1EtX1P044moYZPfkCpsKRVBAA7J2Vtn5g@mail.gmail.com","threadId":"28773","inReplyTo":"CAFfmPPNK+D=g5h7bdYmON++HE5jF_opYxKLobqjOosj--8+9FQ@mail.gmail.com","subject":"Re: FYI: status of svn-fe speed","fromName":"Tay Ray Chuan","fromEmail":"rctay89@gmail.com","sentAt":"2011-10-26T16:18:13Z","receivedAt":"2011-10-26T16:18:13Z","isPatch":false,"sender":{"key":"rctay89@gmail.com","avatar":"https://avatars.githubusercontent.com/u/61553?v=4"},"body":"On Wed, Oct 26, 2011 at 11:18 PM, David Michael Barr\n<davidbarr@google.com> wrote:\n> I talked a lot about low-level speed optimisation after the high-level\n> optimisation is finished this week. I merged my svn-fe-pu branch with\n> jch/next and tested with a 1000 rev dump of patches from git-core.\n> When not blocked, svn-fe can process such revs at ~5000 rev/s on my\n\nI had to rub my eyes when I read this - what???\n\nGood job, then!\n\n-- \nCheers,\nRay Chuan\n"}]}