Dear Aurélien, thanks for the detailed analysis! Sorry, Outlook Web App doesn't seem to support proper quoting of E-Mails, replied mail is just appended... My response starts with ">>>>". we tried to resolve \fsgroup4\group in case it is a dfs root somewhere else but it's not, continue rest of mount. ~2 seconds pass here, I'm guessing you waited and ran a command or something >>>> Yes, as I wrote in the original mail the first listing of that directory worked (apparently leading to the majority of log entries), it was the second time that it hung. Could have been 2s until I restarted the list command... Interestingly it's not always the second time but often the very first time I try to list the directory that it hangs, so I'm not sure whether the "already have inode" path is necessary. Also note that DFS access with this very same server and directory has worked absolutely solidly for years and still does using kernel 4.17.19, but not with 4.18 kernels. I'll send logs with 4.17.19 then (may be Tuesday because I'm away from work), however I fear that this will just show page after page of "everything's as usual" messages. Anyway, if I understood you correctly, we want to see whether 4.18 just reacts differently to an error condition that's always present, right? Must be a quite graceful error handling in 4.17 and previous then, since I never missed a single byte in transit... Best regards, Bernd
