Just a note to anyone noticing their docker image size growing leaps and bounds.  Found this lovely thread featuring the Calibre developer and his flavor-full short-sightedness regarding an issue which is almost inevitable with Docker or image installs https://www.mobileread.com/forums/showthread.php?t=278484

 

I had 335GB in my *\appdata\calibre\.local\share\Trash directory.  Created a simple userscript to clean it out daily with a simple rm -rf

Edited February 11, 20224 yr by Phuriousgeorge

Just updated my Calibre docker and I no longer be able to login after entering login and password. It displays:

connecting to sesman ip 127.0.0.1 port 3350
sesman connect ok
sending login info to session manager, please wait...
login failed for display 0

 

Anyone experienced this behavior?

I attached the log file here.

Calibre_log.txt

My docker env disk seems to be growing and looking at the docker container sizes the calibre container is 5.71GB with 4.43GB writeable. This is twice as large as any other container and it seems to be far bigger that I thought it should be. 

 

What size should i expect the container to be and what settings should I check to trim it down (if possible)

Thanks

 

Container size output in spoiler

 

Spoiler

 

1 hour ago, Sivivatu said:

the calibre container is 5.71GB with 4.43GB writeable.

Inside the calibre app settings, not the container config, what path is pointed to your library?

48 minutes ago, JonathanM said:

Inside the calibre app settings, not the container config, what path is pointed to your library?

im not entirely sure where to find that setting, the only place i can find is when i check the new library it says the current library is located at /books which is where i have mapped my media/books folder

 

Giving 'Calibre' another go here.

 

Container is running and I'm able to access the WebUI with out problems.

 

But I'm seen errors on the log and WebUI shows  'Guacamole Client' as tittle.

 

any advice on solving these errors/issue i would appreciated or can these errors be ignored?

 

 

Edited March 28, 20224 yr by Tolete

On 3/10/2022 at 3:34 PM, bthoven said:

Just updated my Calibre docker and I no longer be able to login after entering login and password. It displays:

connecting to sesman ip 127.0.0.1 port 3350
sesman connect ok
sending login info to session manager, please wait...
login failed for display 0

 

Anyone experienced this behavior?

I attached the log file here.

Calibre_log.txtUnavailable

 

Change MaxSessions=50 to 100 in /etc/xrdp/sesman.ini， looks like a buggy version of xrdp was used.

2 hours ago, taoye said:

 

Change MaxSessions=50 to 100 in /etc/xrdp/sesman.ini， looks like a buggy version of xrdp was used.

Thanks. It was my fault. I forgot the login must only be 'abc' as stated on the git page. It is working if I login with abc/mypassword.

Is it possible to assign my own login name? I tried USER and USERNAME variable on the docker, but it didn't work.

Hi everyone, I've been using this docker container for a couple of years.  Yesterday I noticed that something was using 100% of my server's CPU constantly.  I used htop to check the running processes and found that the xmrig miner was running on my server.  I tracked it to this docker.  Inside my calibre appdata folder is a folder named c3pool and inside that is the xmrig miner files.  Has anyone else had this happen?  Is there an issue with the docker image itself or possibly just something isolated to me.  I did have a port opened to access the calibre UI over the internet but I have since closed that.

1 minute ago, cadarn07 said:

found that the xmrig miner was running on my server.

I'd be absolutely shocked if LSIO bundled xmrig with any of their containers.

 

It was because you opened the port over the internet to access the UI, at which point a script kiddie just gained access rather easily, especially if you have no passwords set.  You should use a reverse proxy or something like Wireguard or the like if you need to access the UI's.  Then there is security.

1 minute ago, Squid said:

I'd be absolutely shocked if LSIO bundled xmrig with any of their containers.

 

It was because you opened the port over the internet to access the UI, at which point a script kiddie just gained access rather easily, especially if you have no passwords set.  You should use a reverse proxy or something like Wireguard or the like if you need to access the UI's.  Then there is security.

 

Thanks for the reply - I didn't think this dev would have bundled it either, but I had to ask to be sure I didn't need to switch to a different docker.  I did have authentication set up but I guess that wasn't enough.  I don't really need off-network access to calibre so I'll just leave the ports closed for now.  I'm also uninstalling it, deleting the calibre appdata folder, and reinstalling to ensure xmrig is gone.  If there's anything else I need to check for that you can think of, please let me know.  I checked my other docker appdata folders and didn't see any evidence of xmrig and I'll monitor Unraid's processes over the next several days to see if anything reappears.

Is it normal for the Calibre doctor to indicate on the Unraid Dashboard Screen that there is an active stream? 

 

Last night I noticed my server was running sluggish so I pulled up the webui.  I noticed that my CPU usage on the server was 100% and not moving.  Trying to figure out what was going on, I saw that my eBooks share had an active stream.  I wasn't in there, and the only other person I gave access to was my brother so he could access my books.  I texted him and he wasn't even home.  UH OH!  This immediately put me into panic mode.  I disabled the docker (which throws up an execution error btw).  That definitely fixed the 100% usage.  After that was done, I saw the active stream disappear.  I then went to edit the docker config and gave the GUAC_USER and GUAC_PASS a completely random login rather than the generic `abc`. After that, I re-enabled the docker.  The webui never prompted for the username/password though, which I thought was odd.  Also, almost immediately, the active stream appeared again.  This now has me worried that some hacker has managed to gain access to that share and figured out the username/password to one of my remote access accounts.  So I went into the Calibre server settings and removed my brother's account and changed the username/password to my account.  Shut down and restarted the docker.. it STILL shows an active stream.  So what gives?  Is this normal/typical behavior?  I posted about this in the Container sub-channel on the Unraid Discord last night but nobody responded.  I'm really concerned here.  If this is normal, that's fine, but if it's not..what do I do?!

 

https://discord.com/channels/216281096667529216/780512072517025823/978527053538295818

On 4/21/2022 at 3:31 PM, taoye said:

 

Change MaxSessions=50 to 100 in /etc/xrdp/sesman.ini， looks like a buggy version of xrdp was used.

 

Thanks for this - worked for me. Hopefully this gets updated in the next release so that we don't have to keep changing this with every update.

 

Cheers

Hi all

 

I literally just updated my Calibre and now it doesn't launch - The docker starts up normally and doesn't seem to generate errors...but when I try to go to the WebGUI, I just get a black screen in the browser window.   The little circle appears that lets me choose to show the on-screen keyboard and browser files etc is there and works, but Calibre never launches and, since the move to Guacamole, if I don't launch Calibre at least once in a browser window...the OPDS server doesn't start and I can't connect my apps.

 

Would anyone have any ideas?

 

Thanks

1 hour ago, PureLoneWolf said:

I literally just updated my Calibre and now it doesn't launch

 

Would anyone have any ideas?

 

Had the same problem today. But right before I sent my original post (just now) it found a new update and it seems to be working again

Edited July 11, 20223 yr by Vipa

19 minutes ago, Vipa said:

 

Had the same problem today. But right before I sent my original post (just now) it found a new update and it seems to be working again

Good shout - Thank you   I had checked a few times after it didn't work - But hadn't checked in a while.  All up and running again now  

Hi,

I'm also having the container size a little too big to be normal (the calibre-web too to be honest), I checked the path and library mapping and i don't see anything strange.

Am I doing something wrong?

the /mnt/user/data/media/Libri by itself is 958 MB

 

Spoiler

Hi all

 

I just updated my Calibre and I can't seem to access the guac client - Just getting server refused to connect - I didn't change anything, just updated.  I can't see any errors in the logs at all - Anyone else having issues?

1 hour ago, PureLoneWolf said:

Hi all

 

I just updated my Calibre and I can't seem to access the guac client - Just getting server refused to connect - I didn't change anything, just updated.  I can't see any errors in the logs at all - Anyone else having issues?

 

I have exactly the same issue. I don't see errors in the logs, but it just refuses to connect. Haven't changed any settings, nor any firewall rules. Any suggestions are much appreciated. 

I just rolled back to 6.4.0 and i am able to connect. Before i was getting

error launching startup command failed to close file

 

6 hours ago, rh535 said:

I just rolled back to 6.4.0 and i am able to connect. Before i was getting

error launching startup command failed to close file

 

Same issue as you and followed your suggestion to rollback to 6.4.0 and it works again.  Something is wrong with the 6.5.0 container.

I'm having the same problem.  How do I roll back? 

 

1 minute ago, hansolo77 said:

I'm having the same problem.  How do I roll back? 

 

Edit your docker and change the repository to:

 

linuxserver/calibre:6.4.0

I can confirm that 6.5.0 is broken for me with the same error, and rolling back to 6.4.0 fixed it. Don't see anything obvious about this on the linuxserver.io github page.

Thank you guys for posting here as soon as you realized there was an issue.  Had a few updates break some things this week, and you guys just saved me some headaches after dealing with everything else.  Rolling back now!