Slow login

[joe_schmo]

I have an RN316, a 428, and an Ultra4.

On the 428 and the Ultra4, when I SSH in via Putty, I put i nmy username and then immediately, the password prompt appears.

With the RN316, it takes up to 30 seconds for it to appear. 

Does anyone know what could cause the system to be slow like this?

[StephenB]

Do you have disk spindown enabled on the RN316?

[joe_schmo]

Nope... although that would have been awesome if it were the case

[StephenB]

Once you do get in, perhaps run top and see if something is hogging the CPU.  Maybe also check the disk health, and the fullness of the OS partition.

[joe_schmo]

root@NAS:~# btrfs fi us /data | grep unall
    Device unallocated:            4.39TiB

root@NAS:~# btrfs balance status -v /data
Balance on '/data' is running
560 out of about 4334 chunks balanced (19676 considered),  87% left
Dumping filters: flags 0x7, state 0x1, force is off
  DATA (flags 0x2): balancing, usage=78
  METADATA (flags 0x2): balancing, usage=39
  SYSTEM (flags 0x2): balancing, usage=39

root@NAS:~# df -h
Filesystem      Size  Used Avail Use% Mounted on
udev             10M  4.0K   10M   1% /dev
/dev/md0        4.0G  3.1G  779M  81% /
tmpfs           992M     0  992M   0% /dev/shm
tmpfs           992M  5.3M  987M   1% /run
tmpfs           496M   11M  485M   3% /run/lock
tmpfs           992M     0  992M   0% /sys/fs/cgroup
/dev/md127       37T   29T  7.4T  80% /data
/dev/md127       37T   29T  7.4T  80% /home
/dev/md127       37T   29T  7.4T  80% /apps

After 13 days, it's at 87% left on the balance.  Any ideas?

[StephenB]

Did you start this balance from the web ui, or from ssh?  If the latter, do you recall what parameters you used?

Also, your OS partition is very full.  What have you installed on it?

[Danthelf]

Do you have the NAS joined to AD or did you ever have it joined to AD? I never really investigated it further but my NAS is doing the same and I seem to recall it started happening after I joined the NAS to AD, despite having now switched back to 'Local Users'.

Don't know if I'm just imagining things but at least I think that's when it started.

[Danthelf]

For the sake of it I decided to look in to this further, so in order to find out what was going on I did a packet capture while logging in. The reason for the slowness is because the SSH server will attempt to do a reverse DNS lookup the hostname of the client you're connecting from before presenting you with the password prompt, so if your DNS settings are misconfigured it will have to wait for that to timeout before presenting you with the password prompt.

The reason why I found this to be related to me joining the NAS to my AD domain is simply because at the time I modified my DNS settings to use:

1) My DC's IP

2) google DNS

My DC is currently turned off (this is a home lab) so of course it will not respond to DNS queries. Removing it from the list of DNS servers resolved the issue and logins are now instant.

You can also stop the ssh server from attempting lookups by adding "UseDNS no" to your sshd_config

# echo "UseDNS no" >> /etc/ssh/sshd_config
# systemctl restart ssh

[StephenB]

Thx for posting the root cause and solution to your problem 

[joe_schmo]

This is interesting... but would that explain the slow balance and the slow rsyncs from it as well?  

[joe_schmo]

WebUI... here are the parameters:

DATA (flags 0x2): balancing, usage=78
  METADATA (flags 0x2): balancing, usage=39
  SYSTEM (flags 0x2): balancing, usage=39

[StephenB]

---

@joe_schmo wrote:

would that explain the slow balance and the slow rsyncs from it as well?  

No it doesn't.

---

@joe_schmo wrote:

WebUI... 

When was the last time you ran a balance?

What firmware are you running? Netgear made a change a while back to change the usage threshold (not sure exactly when).

Overall, the usage threshold is defined here: https://btrfs.wiki.kernel.org/index.php/Manpage/btrfs-balance  The higher the threshold percentage, the more chunks are re-written.