Re: Apparent timeouts
Olaf Hering <[email protected]>
| Newsgroups | gmane.network.unison.general |
|---|---|
| Message-ID | <[email protected]> |
Sun, 21 Jan 2024 18:31:57 +0100 John Rose <[email protected]>: > Is there a set-up parameter to increase the timeout threshold or something else to eliminate or reduce this problem? There is a knob in /proc or /sys which controls when the kernel starts to write accumulated dirty data to a block device. For potentially slow and/or unresponsive devices that value should be low. The built-in kernel (or Ubuntu) defaults are likely not suitable for the involved hardware configuration. While a process tries to create more dirty data, it has to wait until the existing dirty data have been flushed. The unison process is likely in the D state when the window manager complains about an unresponsive application. The only solution here is to have a separate IO thread, which processes the IO, while the UI thread processes window manager events. Olaf -- To unsubscribe from this group and stop receiving emails from it, send an email to [email protected].
signature.asc
(application/pgp-signature, 833 B)
-----BEGIN PGP SIGNATURE----- iQIzBAEBCAAdFiEE97o7Um30LT3B+5b/86SN7mm1DoAFAmWthm4ACgkQ86SN7mm1 DoATDxAAlsr2dsDUgJ9hrIeMw203zZ0daou8ko6Artai8Y5lxgsEps5b3TLAPmSt Uev0Fw/wsYMAmu4VB7OdWuwNbHqbr/j0p5RU9OlndnROJpUZHe8gnnHKiVmIe+Jm cMCmSAjwYbTDMWpIXo4hLxtZrXZOmti13SBWVcaiLchoaGdfinLa14UyG+auQesC 9EKOZ19q9igj0vC9KaN3vqzWhdYFKK0ununzyY2vsPW+Bt76twHsyl6UoZDHssSt kEs1sJG8mPGS5E4/uUkueWh8pm1rz4NHWZFIRKrLuZaJc0oxmJzpug2P2fLxSnBk yRQT+o2LDpEdGTCTexnw1yQITy4G0sWezcsNYrwal+GROu4FBsFwY1JHTlqpbi4b cVGyPLXQnmLS2NnsqwQ1RQsx9KvJA0T6gcHOLzDz4SUCU7s65dvjWGWw9hY58WV7 5tBhj6+xWFkOMICGvLLgDtqQuDcgnRG/27Zs2I/0GYUQKRgO1HQd444iGLQvykG9 xQ6Ay6CUODaCVPQ0LBhqslb6EgBO2HCvfcOKxuPyMnsro/D0wddcaba30OS1Sr2T yIzeoUQ9tCVpJsAod+X+4v9gKQeou7PRwq+xF6wFpwWgQwEtIKssTkWp2T685d9T pM/H3BVkHfv4mBDzqBXtn+YcFeddDMITpd9OYk6U+vTPS0FyS+8= =Rtyf -----END PGP SIGNATURE-----