-----BEGIN PGP SIGNED MESSAGE----- Hash: SHA512 I'm staying, but I need some help.
My server (sks.daylightpirates.org) currently has an uptime of 97.95%, but every few hours the "sks db" process spikes to 100% CPU which makes the web interface completely unresponsive, which I'm guessing is why I keep getting kicked out of the pool. In the db.log, for each of these events, there's the following logs: ================ 2018-07-19 18:04:30 add_keys_merge failed: Eventloop.SigAlarm 2018-07-19 18:04:38 Key addition failed: Eventloop.SigAlarm ================ After "Key addition failed", my server becomes responsive again. I'm not having storage problems (/var/lib/sks is only 15GB, probably due to DB_LOG_AUTOREMOVE being set), but the CPU spikes are really annoying since it kicks me out of the pool. Is there anything I can do to prevent these CPU spikes or offload them to another process so the web interface stays up? Thanks! Daniel Roesler diaf...@gmail.com For reference, here's my sksconfig: ================ # /var/lib/sks/sksconf debuglevel: 3 hostname: sks.daylightpirates.org nodename: server hkp_address: 0.0.0.0 hkp_port: 11371 recon_port: 11370 server_contact: 0xE7F6FAD172EFEE3D initial_stat: disable_mailsync: membership_reload_interval: 1 stat_hour: 12 max_matches: 500 ================ And my KDB/DB_CONFIG: ================ set_flags DB_LOG_AUTOREMOVE ================ -----BEGIN PGP SIGNATURE----- Version: GnuPG v1 iQEcBAEBCgAGBQJbUNUEAAoJEMtwcDcM6wt6UBAH/14emyqhSmZIOs0ZPMPvvy56 wemg2BHS5VO4zqJft8OZKbd5q5Oj4DwFI56IXMpLEhDpFsg8yqHJ8YMzXeIP3g5+ y5oNKcGbPwYTcZLfyKZbx2IoK4az9Bau+soDO4U2DZv35SUPfEYTmL03XZa2gVTA Ik6huxqjQv1Vo+llC20nCmHhj0e1OlZhAYmPB3WY7mdPahQ4KFtO/c8y/Cpzs1H7 85WSQoTBb9KEkR0KSDyaqxunW2h9t6MCcpwkglllRYwledlL7bvtslMcHJ1fS9z7 NZatjJuzj1WvMuM6t//FNoHyLRQ3NeA3kUds+k4WtjLE9iAlRJZa6zTUw4dbUUA= =zg7A -----END PGP SIGNATURE----- _______________________________________________ Sks-devel mailing list Sks-devel@nongnu.org https://lists.nongnu.org/mailman/listinfo/sks-devel