2017-08-28 07:01 AM - edited 2017-08-28 07:04 AM
With RHEL 7.4 being released recently, we proceeded with upgrading a decent number of servers to that rev.
With 7.3 having 4.0 as the default version and 7.4 having 4.1, we noticed a problem with a few of the servers we upgraded afterwards.
I won't get too descriptive with what we saw but it affected some Wordpress instances that we are serving up from an NFS mount residing on our FAS8200. The webpages would come up very slowly or not at all. I/O wait time spiked up really high. From a terminal session, 'ps -ef' command would hang up. Rebooting the server would take a very long time (10+ minutes). All a result of mounting NFSv4.1 versus it being NFSv4.0 prior.
We've been leveraging NFSv4.1 with autofs homedirs for some time now and not noticed anything like this.
Anyone else seen something like this? Would you suspect this to be something with our configuration on the linux client or something in the NFS configuration on the SVM?