ChrisReeve
Explorer
- Joined
- Feb 21, 2019
- Messages
- 91
Hi
I have recently built a server and installed freeNAS and Plex. Those are the two main uses. I have a jottacloud.com account which provides unlimited storage for backup, and I am trying to back up my media library (among other things). Approx 10TB total size.
My problem is that my entire freeNAS server (more or less) freezes. Upload-speeds drops from approx 60MB/s to 0-2MB/s, and the NAS becomes inaccessible from my computer until I restart the server.
When I try to shut down via the freeNAS webGUI, nothing happens. I then try to shut down via IPMI webGUI. After a few minutes, I get the following error message on the server screen (still running):
"freenas init: timeout expired for /bin/sh on /etc/rc.shutdown: Interrupred system call: going to single user node"
The only way to shut down is to cut the power to the system, or hold the power button for 5 seconds (hard shutdown).
My setup is the following:
Supermicro X9SRL-F
Xeon E5-2630L v2
24 GB Actica memory (6x4GB DIMMs, ACT4GHR72Q4G1333H)
1x Intel 180GB SSD for boot
9x WD RED 4TB NAS drives in one ZFS2 pool
Temperatures are good. CPU is usually 30-45C, with RAM being a little higher during peak loads, reaching up to 65C (30-40C during low loads)
I had 8x4GB, 32GB total, but removed two DIMMs installed in slot C1 and C2, as i got constant error corrections and slowdowns from slot C1, and sometimes from C2, reported in IPMI event log. Still troubleshooting this, but after removing those, the server has been rock stable, until now...
I do belive my current problems are separate from the RAM issues, as removing slot C1 and C2 have fixed the problem, and the symptoms were different.
I am connected via 1Gb ethernet locally, on a 500mbit symmetrical fiber internet connection. Read speeds vary from 60-100MB/s on the server (reported from freeNAS). Writes are close to zero during backups.
The workflow is as follows: I have installed the Jottacloud client on my main computer (Windows 10, newest version). I have selected the folder(s) on my connected NAS to back up in the Jottacloud-client, and it starts checkuming and uploading the files. Therefore, my main computer accesses files over ethernet, then uploads them to jottacloud.
On reboot, the server boots normaly, and works fine. Until I try to backup for more than a short amount of time.
I also do not belive it is caused by any bugs in the Jottacloud-software, as I would assume I will be able to replicate this freeze by any high IO-load on the server/pool. If you disagree, I am happy to do tests, but unsure how to do this.
I have of course tried to google my problem, and many users are experiencing similar issues while pushing a high IO-load on their servers, and claim there is a known problem with the ZFS file system. What do you guys think?
I have recently built a server and installed freeNAS and Plex. Those are the two main uses. I have a jottacloud.com account which provides unlimited storage for backup, and I am trying to back up my media library (among other things). Approx 10TB total size.
My problem is that my entire freeNAS server (more or less) freezes. Upload-speeds drops from approx 60MB/s to 0-2MB/s, and the NAS becomes inaccessible from my computer until I restart the server.
When I try to shut down via the freeNAS webGUI, nothing happens. I then try to shut down via IPMI webGUI. After a few minutes, I get the following error message on the server screen (still running):
"freenas init: timeout expired for /bin/sh on /etc/rc.shutdown: Interrupred system call: going to single user node"
The only way to shut down is to cut the power to the system, or hold the power button for 5 seconds (hard shutdown).
My setup is the following:
Supermicro X9SRL-F
Xeon E5-2630L v2
24 GB Actica memory (6x4GB DIMMs, ACT4GHR72Q4G1333H)
1x Intel 180GB SSD for boot
9x WD RED 4TB NAS drives in one ZFS2 pool
Temperatures are good. CPU is usually 30-45C, with RAM being a little higher during peak loads, reaching up to 65C (30-40C during low loads)
I had 8x4GB, 32GB total, but removed two DIMMs installed in slot C1 and C2, as i got constant error corrections and slowdowns from slot C1, and sometimes from C2, reported in IPMI event log. Still troubleshooting this, but after removing those, the server has been rock stable, until now...
I do belive my current problems are separate from the RAM issues, as removing slot C1 and C2 have fixed the problem, and the symptoms were different.
I am connected via 1Gb ethernet locally, on a 500mbit symmetrical fiber internet connection. Read speeds vary from 60-100MB/s on the server (reported from freeNAS). Writes are close to zero during backups.
The workflow is as follows: I have installed the Jottacloud client on my main computer (Windows 10, newest version). I have selected the folder(s) on my connected NAS to back up in the Jottacloud-client, and it starts checkuming and uploading the files. Therefore, my main computer accesses files over ethernet, then uploads them to jottacloud.
On reboot, the server boots normaly, and works fine. Until I try to backup for more than a short amount of time.
I also do not belive it is caused by any bugs in the Jottacloud-software, as I would assume I will be able to replicate this freeze by any high IO-load on the server/pool. If you disagree, I am happy to do tests, but unsure how to do this.
I have of course tried to google my problem, and many users are experiencing similar issues while pushing a high IO-load on their servers, and claim there is a known problem with the ZFS file system. What do you guys think?
Last edited:
