Skip to main content

SSH Troubleshooting Information

Connection Timed Out

The SSH connection will timeout after you have had multiple failed login attempts. This is a soft ban that only lasts for 10 minutes. In this case, please change your SSH password from the Control Panel and wait for 10 minutes. Then, try to log in via SSH using the newly set password. Attempting to log in even with the correct password when you are soft-banned will just get your IP banned again for 10 minutes. Therefore, it is important to wait patiently for 10 minutes.

How to check high disk IO utilization?

If your apps are slow and your slot is running sluggish, this is the most common cause.

Any file access on the disk creates io load on the HDD, either by you or other users in the node. This is not an issue with the NVME slots due to their breakneck IO speeds.

Typical applications that do this include the following:

  • Torrent clients "checking" torrents.
  • A very active torrent.
  • Unpacking files
  • Constant rclone operations to and from providers
  • Usenet operations (Unpacking, Downloading, Repairing, etc.)

Execute this command in your shell terminal and pay attention to the last column %util.

iostat -xk 2 $(findmnt -T ~ | awk 'END {print $2}')

If this is 100% after a few minutes, the disk is entirely saturated, affecting performance. We recommend reviewing the settings in your installed applications to prevent this, e.g. limiting concurrent activities by utilizing queuing when available. See the following section to find what application is causing high disk io.

Check IO utilization per individual process:

If you suspect that you are the reason for the high IO utilization rather than one of your disk neighbors, you must find out which process is the culprit. There are two ways you can investigate this. The preferred way is to use htop.

  • Execute htop from your shell
  • Maximize your ssh client to full screen. If you don't see a column called DISK R/W, follow these steps:
    • Press F2 for access to settings
    • Navigate to Columns
    • Add IO_RATE
    • Press Escape to exit settings

You should now see the newly addedThe DISK R/W column whenwill you viewshow your process list in htop. Take note of thedisk IO transferrate rates,per andindividual thenprocess. exit htop by pressing F10 to save your new column selection for next time.

Tip: Clicking on the column headings toggles between ascending and descending sort order for the selected column.

See which process has the highest DISK R/W value, paying attention to whether it is B/s, K/s or M/s which is bytes, kilobytes, or megabytes per second respectively.

Besides using htop, you can also use pidstat to see the IO utilization per individual process. Execute the following command: pidstat -dl 2

If the above steps do not identify any of your processes as causing high disk io, the issue could still be your FTP activity; please check if your FTP client is downloading multiple streams and/or running without a speed limit. Failing that, another customer hosted on the same disk as you may be causing the high disk io. Please raise a support ticket if the issue continues. The support team will review all customers on the same disk as you.

How to clear up disk space or move data from my slot?

The easiest way to manage any data is by using your favorite S/FTP client. Filezilla is the most popular choice. We recommend using SFTP as your connection of choice because it is more flexible on which folders to manage. You can also drag and drop with an S/FTP client, which makes it incredibly useful as a whole. The downside of this is that it takes time for the client to process your moves. Our help guide for S/FTP is available here.

However, the fastest way is by using SSH, but this might be intimidating for inexperienced users. For starters, you can use Midnight Commander, which is a graphical user interface file manager in your slot. It can be started by executing mc in your shell. You can also read our basic Linux cheat sheet here.

Why am I unable to connect with SSH, and why is there an error "Resource Temporarily Unavailable"?

In most cases, this is your total proclimit being saturated. We implement resource limits in each slot to ensure fair resource usage on all users. Usually, apps like rclone or syncthing are responsible due to hundreds of worker threads spawned by the application. In most cases, restarting your applications before logging into SSH again should fix the issue. If it does not work, you may need to submit a ticket.

Some of my files have duplicates in ~/files or ~/Downloads directory and ~/media directory. Are they using double disk space?

If you're using Radarr/Sonarr/Sickchill/CouchPotato/Medusa etc., they should not be taking any additional space. By default, these applications create a hardlink of the file to the default root directory of your application. A hard link can be described as a single file located at two different paths but pointing at the same block of data on the disk. Think of it as two signposts from other streets that direct you to the same destination. Hardlinks do not take any extra space. You can compare the inode numbers for both files from the Terminal (SSH Access) with ls -I on both files'. The inode number will be the same.

You can read more about Hard Links here: https://en.wikipedia.org/wiki/Hardlink.

Back to top