вт, 17 сент. 2019 г. в 23:49, ronnie sahlberg <ronniesahlberg@xxxxxxxxx>: > > On Wed, Sep 18, 2019 at 4:16 PM James Wettenhall > <james.wettenhall@xxxxxxxxxx> wrote: > > > > Thanks Pavel, > > > > We've been running Kernel v5.2.14 over the past week (updated using > > Ukuu) and it seems to have improved the situation considerably. > > Thank you for the feedback. > This is very good news. > > > > > > I assume that the "nohandlecache" mount option recommendation was just for v5.0. Glad to know that the situation have improved for your workload. Thanks for the feedback. v5.2 kernel has many fixes preventing reconnects that's probably why you stopped observing the original problem. The latter haven't been completely fixed yet in v5.2.y yet. We have a patch in for-next that aims to fix it but it haven't been sent to the mainline yet, see https://git.samba.org/?p=sfrench/cifs-2.6.git;a=commitdiff;h=96d9f7ed00b86104bf03adeffc8980897e9694ab. Once it is there, it should be automatically picked up for backporting to all active stable kernels it applies to. In the meantime, If you start hitting the issue again, please try mount option "nohandlecache" as a workaround. -- Best regards, Pavel Shilovsky
