Dear Bruce, thank you for quick response. And yes - it's no an easy one for sure, therefore i really need your kernel gurus expertise! We have no contract for Linux, it's based on Gentoo Linux at all: We are totaly free to try out anything you ask for. This "datacenter design" now is working for about 8 years using a whole bunch of kernel versions without any problem. The issue *may* have start to appear in 2018Q1, maybe with with changing to LTS 4.14 or with changes concerning the Spectre theme. It have happended two/tree times on different other container hosts this year acting in Test and Approval stage. Because of the shared NFS infrastructure, all using exactly the same kernel image and (template-sourced copies of) the same root image. This also some older rackservers (for Evaluation stage) with comparable smaller hardware. They just have "1GB"-NICs and there might be a clue that the issue may be forced there by heavy file IO workload. I have to re-check this. Unfortunately, the Apache email infrastructure is problematic (don't accept some mail encodings), but in the end I was able to create an account and open an issue (https://bugzilla.linux-nfs.org/show_bug.cgi?id=328). But I still can't attach things to this, i just got some "internal error". Therefore, please ask and I'll send it via email. I have a photo of the "SysRq" console showing a locked task and a tcpdump of "port nfs" taken at the last event. I'm may send kernel config file, kernel image or whatever you need. thank you all in advance Guido Kernel history: root@bladerunner14 ~ # ll /boot{,/_save}/kernel* -t | grep -v old lrwxrwxrwx 1 root root 38 Sep 18 11:46 /boot/kernel -> kernel-genkernel-x86_64-4.14.65-gentoo -rw-r--r-- 1 root root 5.3M Sep 18 11:46 /boot/kernel-genkernel-x86_64-4.14.65-gentoo -rw-r--r-- 1 root root 5.3M Aug 8 12:31 /boot/kernel-genkernel-x86_64-4.14.61-gentoo -rw-r--r-- 1 root root 5.3M May 24 12:58 /boot/kernel-genkernel-x86_64-4.14.43-gentoo -rw-r--r-- 1 root root 5.3M Apr 9 12:20 /boot/kernel-genkernel-x86_64-4.14.32-gentoo -rw-r--r-- 1 root root 4.5M Feb 27 2018 /boot/_save/kernel-genkernel-x86_64-4.9.84-gentoo -rw-r--r-- 1 root root 4.5M Jan 23 2018 /boot/_save/kernel-genkernel-x86_64-4.9.76-gentoo-r1 -rw-r--r-- 1 root root 4.5M Nov 16 2017 /boot/_save/kernel-genkernel-x86_64-4.9.61-gentoo -rw-r--r-- 1 root root 4.2M Jan 3 2017 /boot/_save/kernel-genkernel-x86_64-4.4.39-gentoo -rw-r--r-- 1 root root 4.0M Oct 6 2015 /boot/_save/kernel-genkernel-x86_64-3.14.51-gentoo -rw-r--r-- 1 root root 4.0M Aug 28 2015 /boot/_save/kernel-genkernel-x86_64-3.14.9-gentoo -rw-r--r-- 1 root root 3.7M Jul 14 2014 /boot/_save/kernel-genkernel-x86_64-3.10.20-gentoo -rw-r--r-- 1 root root 3.5M Oct 10 2013 /boot/_save/kernel-genkernel-x86_64-3.8.13-gentoo -rw-r--r-- 1 root root 3.4M Apr 23 2013 /boot/_save/kernel-genkernel-x86_64-3.3.5-gentoo -rw-r--r-- 1 root root 3.3M May 16 2012 /boot/_save/kernel-genkernel-x86_64-3.3.4-gentoo -rw-r--r-- 1 root root 2.7M Dec 1 2010 /boot/_save/kernel-genkernel-x86_64-2.6.34-gentoo-r6 On 19.09.2018 21:13, 'J. Bruce Fields' wrote: > That just looks hard to debug, unfortunately. Have you tried asking > Netapp, or do you have a support contract for your Linux clients? Was > there an older kernel that worked OK? >
