+6
Under review

Portal not responding

frufru 1 month ago updated by Tns 1 week ago 23

Today I tried to install the latest version and the problem is still there. I start the service and everything works fine. After an hour I can’t connect to the portal, sometimes half of the page loads but can’t see any clients. While this happens I see high cpu usage from multiple mono instances in top. If I restart the service it works again for an hour. Now back to 19.4 for the hundredth time..

Thanks for letting us know, we have received several reports and are investigating.

Can you post the last 100 lines or so from /var/log/screenconnect ?

Also, what Linux OS is on the server?

Hi,

Here is the relevant part of the log. The first entry is the service stop before the upgrade. The system stopped responding after about an hour but no sign of the problem here.

Event (2020/12/08 10:46:06.900 +01:00, ScreenConnect Web Server, Information): Stopping service...

Event (2020/12/08 10:46:07.038 +01:00, ScreenConnect Web Server, Information): Stopped service.

Event (2020/12/08 10:46:07.039 +01:00, ScreenConnect Relay, Information): Stopping service...

Sending SIGKILL.

Event (2020/12/08 10:56:19.264 +01:00, ScreenConnect Session Manager, Information): Starting service...

Event (2020/12/08 10:56:20.264 +01:00, ScreenConnect Session Manager, Information): Successfully started service.

Event (2020/12/08 10:56:20.264 +01:00, ScreenConnect Relay, Information): Starting service...

Event (2020/12/08 10:56:20.862 +01:00, ScreenConnect Relay, Information): Successfully started service.

Event (2020/12/08 10:56:20.862 +01:00, ScreenConnect Web Server, Information): Starting service...

Event (2020/12/08 10:56:51.700 +01:00, ScreenConnect Web Server, Information): Successfully started service.


Interestingly today I left the server without a service restart after it hang and after some time it has recovered and at this moment I can access it again but I don’t know for how long.

One more thing I experienced, the upgrade from 19.4 fails with the following error, I need an intermediate step:


The assembly mscorlib.dll was not found or could not be loaded.

It should have been installed in the `/root/monoInstall/lib/mono/4.5/mscorlib.dll' directory.

License is not eligible to be upgraded to this version

it hang again, after a couple of hours, here is a top screenshot:

there is nothing in the log

It does this even when the VM is isolated and no clients are trying to connect to it.  This is running under KVM/QEMU

Yeah I was having this problem so I setup a fresh VM on another computer and it is also having the problem.  The VM is running Ubuntu 20.10

It is worth mentioning that I did adjust for the libc bug you are having when I set this up by symlinking to the files your app needs but is not finding.

(If you gave us a working docker image that is a much more controlled environment and would be much easier for you to manage, debug, and release updates for)

We had the same problem since 20.1.29489.7513 under debian 9 stretch

I can't say if this fixed the problem permanently, but currently (Since the latest Debian updates yesterday) it seems to be gone.

With the updated packages, I can't really see a connection to screenconnect now either.
But maybe your technicians see a connection here ?
Here are the packages that I updated yesterday:


Start-Date: 2020-12-10 15:49:12

Commandline: apt upgrade

Upgrade: libapt-inst2.0:amd64 (1.4.10, 1.4.11), libldap-2.4-2:amd64 (2.4.44+dfsg-5+deb9u5, 2.4.44+dfsg-5+deb9u6), apt:amd64 (1.4.10, 1.4.11), libsqlite3-0:amd64 (3.16.2-5+deb9u2, 3.16.2-5+deb9u3), python-apt-common:amd64 (1.4.1, 1.4.2), libapt-pkg5.0:amd64 (1.4.10, 1.4.11), libldap-common:amd64 (2.4.44+dfsg-5+deb9u5, 2.4.44+dfsg-5+deb9u6), python-apt:amd64 (1.4.1, 1.4.2), apt-utils:amd64 (1.4.10, 1.4.11), apt-transport-https:amd64 (1.4.10, 1.4.11), python3-apt:amd64 (1.4.1, 1.4.2)

End-Date: 2020-12-10 15:49:16

I can safely confirm that the service has been running continuously without any downtime because since the problems with 20.1 we have been monitoring it with our own monitoring solution ServerSentinel every 10 minutes and as soon as the service was frozen we restarted it via a remote script action so that our developers can use the system when they have a support case.

Unfortunately a day later the problem occurred again.
Possibly this post can be deleted because misleading. It would have been too nice if this problem would have solved itself so easily after three quarters of a year.

+2

I do not understand why this is taking them so long to fix.  I have had to setup a windows computer in my house to compensate for this.  It is a HUGE pain.  They call us "partners" but I really do not feel like one.  Do they have like one programmer working on this linux problem???  I think they changed hands at some point.  After that things started going down hill.  It is a shame too because this software was (sort of is) great.  But the things that made it great are being neglected and fading away.

+1

I ran updates on my Ubuntu VM and this is still a problem. 

+2

What is the update on the resolution for this? Would like to see this resolved before the New Year!

They closed my ticket and asked me to check here for updates. Nothing but crickets.

we should get the hint

You mean they? As far as I'm concerned, they have breached their contract by not fulfilling their duties and obligations. Who is the head of this service corporation anyways?

+2

That is what they did to me too!!!!   

I do not understand why this is taking them so long to fix. I have had to setup a windows computer in my house to compensate for this. It is a HUGE pain. They call us "partners" but I really do not feel like one. Do they have like one programmer working on this linux problem??? I think they changed hands at some point. After that things started going down hill. It is a shame too because this software was (sort of is) great. But the things that made it great are being neglected and fading away.

Can we get an ETA on this issue????

ConnectWise?? Are you actually monitoring this?  Can we get a comment??

+1

So frustrated. I've kept upgrading hoping it will get better, but if anything 20.2 is worse. Restarting the service takes way longer and it stays up for even less time than 20.1. Really wish I could go back to the versions that worked. 

Happy New Year! Where are we at with this?

+1

The “Initial” bug report about this problem has been created almost a year ago (https://control.product.connectwise.com/en/communities/6/topics/2798-portal-not-responding).


Back in December ConnectWise released a security bulletin concerning a vulnerability (https://www.connectwise.com/Company/Trust/Security%20Bulletins#4ae056a7-e060-4b6f-93fd-cd4f3a2f8310) Since the fix is in version 20.13 and the Linux version is way way behind, it seems we are stuck with a vulnerable version!


@caitlin, I think we have been very patient. Like you asked, bug reports have been filled but it seems the resolution of the problem is going at a crawling speed! As a Product Manager for ConnectWise, you shouldn’t let us in the dark by giving us the silent treatment! Please be more pro-active and keep us informed about the progress (if any) and be honest with us!

+1

We've had to set up a cron job to restart the stupid service every hour.. that's the only way the server remains usable.. otherwise it gets slower and slower until finally stops listening to TCP requests for https and dies. I am annoyed as all hell because I was working just fine on an older version for a long time, then macOS clients started acting wiggy with the new macOS release, so I paid to update my license and upgrade to the latest version, and this is what I get..

I feel your pain! I have even gone as far as to re-setup our server on Windows, however we were faced with a dismal performance drop to our other line-of-business applications and went back to Linux. We've been surviving but it's extremely frustrating living in a state of limbo (two if you count COVID) for nearly a year.

+1

I think connectwise removed the linux reports from the main bug portal so this small bunch of annoying people won’t show up all the time and just be in the way with this stupid “portal is not responding” stuff. It looks like no one reads these messages but at least we can talk to each other…:)

I’m lucky because I still have a working 19.4 instance I can go back to, but sometimes I think about the major security issue if my version has is or not.

+1

If you all want them to pay attention to this.  Regularly start new trouble tickets and call sales and demand they give you a further discount on renewing your licenses since they are not following through on their commitment to supply updates.  I plan on demanding they reduce my 20% to 10% for this last year when my renewal comes up.

Echo... echo... echo.... echo.... echo... echo...

Is there anyone from ConnectWise listening here?

I doubt there is which is why you need to keep calling and demand a discount.

I just tried to submit a support ticket (https://www.connectwise.com/services/support) about the security bulletin "ConnectWise Control Host Header Injection". where I wanted to know when a secure version will be available for us, their Linux customers but I get a "403 Forbidden" when clicking the Submit button! What's the best way to contact them beside phone?