Eh, I installed it on both my 6 bay units... so far nothing catastrophic. I can't really find out what they changed (nothing apparent... things seem to work normally)
I picked up the RN214 last week (my first ReadyNAS product) and noticed that readynasd was crashing every night on the 6.4.0 firmware.
Since the web interface was unavailable (httpd was still running, but it obviously couldn't pull data from readynasd) I pulled the logs via RAIDar. It was failing because it indicated not enough RAM.
It tried to relaunch several times, but failed with the same error until it appeared to give up.
Manually running readynasd via SSH worked fine, however after a bit it crashed with a segfault.
I did not update to the first beta of 6.4.1, but did install T29 late last night.
At first all appeared to work as expected, however when I woke up in the morning I was unable to reach the ReadyNAS by any means including ping, had to reboot by pulling the power.
Right now the only app I've installed is Plex using the latest version from their website.
I did grab a copy of the logs from both 6.4.0 and 6.4.1-T29 crashes if they would be beneficial.
updated to 6.4.0 firmware -> crash every night -> cancelled and disabled all the snapshots -> crash every 2/3 days (maybe related to Plex and Transmission apps) -> updated to 6.4.1 T21 firmware -> crash every 2/3 days -> updated to 6.4.1 T29 firmware -> crashed yesterday.
Now I'm supposing that the problem is related to Plex and Transmission apps...
So last night overnight did not hang as it did the previous night, but the management service did still die without recovery.
This time I was able to access with RAIDar and pull logs.
[Sat Nov 7 18:20:28 2015] systemd[1]: readynasd.service holdoff time over, scheduling restart. [Sat Nov 7 18:20:28 2015] systemd[1]: readynasd.service failed to run 'start' task: Cannot allocate memory [Sat Nov 7 18:20:28 2015] systemd[1]: Unit readynasd.service entered failed state.
[Sat Nov 7 18:20:28 2015] Out of memory: Kill process 15813 (Plex Media Scan) score 855 or sacrifice child [Sat Nov 7 18:20:28 2015] Killed process 15813 (Plex Media Scan) total-vm:2357096kB, anon-rss:1817348kB, file-rss:68kB
The Plex overnight processes, such as Media Scan, are scheduled to run 2am to 6am. Does the log timestamping ignore what is configured in the web ui for timezone?
I do have Plex setup to create the "Video Preview Thumbnails" as part of the overnight scheduled maintenance process, which might be contributing to additional load.
A little further down in the dmesg log, entries go from Nov 7 19:20 to Nov 8 1:14 and then back to Nov 7 22:44, followed by the above PMS termination at 18:20. Is it normal for the logs to have entries that are scattered hours in time?
Am I reading it correct that Plex is using 1.8 gigs of the total 2 gigs the RN214 has? Are we able to cap RAM usage per process or does the team at Plex have to do that in code?
It does look like Plex is what is using up the memory.
There is a Plex Media Server app coming for the RN214.
You might want to open a post on the Plex forums if you haven't already and attach the Plex Media Server log to it (there is a different procedure to obtain that log).
