logo
Welcome Guest! To enable all features please Login or Register.

Notification

Icon
Error

Options
Go to last post Go to first unread
CTC  
#1 Posted : Friday, April 21, 2017 9:00:36 PM(UTC)
CTC


Rank: Newbie

Joined: 5/20/2015(UTC)
Posts: 6
United States

Thanks: 1 times
I have ScreenConnect hosted on a Digital Ocean VM. We have about 750 PC's connected and run 6.2.12963.6312.

Previously snappy performance has declined over the last month or so. We're seeing delays of seconds when searching the list of PC's. Remote control sessions have become sluggish. You really notice an issue when typing in a command line on the remote system for example. Mouse activity is click, wait, click, wait... Chat used to be nearly instant but now takes a few seconds. The delays are not quite crippling but they really add up over time.

Performance used to be good on a 1GB RAM single processor VM. I've moved it up two sizes to 4GB RAM and dual processors with no improvement. VM load appears low. It's nowhere near maxing out CPU or RAM utilization.

I don't know if I need to move off of Digital Ocean and bring it back in-house. Or, are there some ScreenConnect maintenance tasks that I should perform? The database maintenance plans are at their defaults of keeping events for 30 days. I've tried compacting database files too.

Any thoughts?

trex2002ro  
#2 Posted : Wednesday, April 26, 2017 7:36:03 AM(UTC)
trex2002ro


Rank: Newbie

Joined: 8/28/2014(UTC)
Posts: 1
Romania
Location: Pitesti

Thanks: 1 times
I have the same issue.
With Edge and Internet explorer searching for an Access PC take 5-10 Seconds. With Chrome is 50% faster but still slow compared with previews versions
vexation  
#3 Posted : Thursday, April 27, 2017 7:59:51 PM(UTC)
vexation


Rank: Member

Medals: Level 1: Random Act of Kindness! Received One Thanks!

Joined: 8/28/2012(UTC)
Posts: 21

Thanks: 4 times
Was thanked: 2 time(s) in 1 post(s)
I've been seeing similar issues, I first noticed it in 6.1 (around the same time we found some unattended sessions continually caused web page errors)

Our "default" session group had all machines in it (2000+), if I leave that session group selected in Chrome and look at the Chrome Task Manager (SHIFT+ESC) it quickly pegs each core at 25% and becomes unresponsive/slow.

We changed the default session group to be empty and our other session groups (averaging 100 machines a group) are typically fine for searching etc.

Have also noticed that since upgrading to 6.0+ the server process on Hosts (i.e if we connect to an unattended session and open task manager) seems to be a lot more hungry than it was in previous version.

It's not the end of the world but figured I'd post in the event it's to be expected or not.
Scott  
#4 Posted : Tuesday, May 2, 2017 10:28:07 AM(UTC)
Scott


Rank: Administration

Medals: Level 4: Wise Old Owl! Received 100 Thanks!

Joined: 3/28/2014(UTC)
Posts: 2,533
United States

Thanks: 3 times
Was thanked: 309 time(s) in 266 post(s)
@all users in this thread, when you say it takes a long time to load, roughly how many sessions are in the session group that you are attempting to view?
ScreenConnect Team
itsbrent2  
#5 Posted : Tuesday, May 2, 2017 1:42:51 PM(UTC)
itsbrent2


Rank: Newbie

Joined: 7/7/2015(UTC)
Posts: 11
United States

Thanks: 2 times
Was thanked: 1 time(s) in 1 post(s)
We upgraded our self-hosted server to 6.2 last weekend and now are experiencing this same performance degradation.

Previously in 5.6 or 6.1 we initially had to wait about 5-10 seconds to load a large group (2000+), but after that we could scroll through the group and find machines quickly with no noticeable performance hit. Now in 6.2 the experience is always slow in these large groups, from load to search to scrolling.

Session Group A has ~1000 sessions and performs fairly well, even if not as snappy performance as smaller groups.

Session Group B has > 2000 sessions and performs very slowly. In Chrome or Firefox, every click takes between 5-10 seconds to reload the page, including clicking the scroll bar, or simply selecting a different session. In IE (no, we generally don't use this) or Edge the page crashes entirely and has to reload, making this group completely unusable in those browsers. While this group is selected, we can't even scroll the Session Groups panel without a significant delay (scrolling only one screen every 2-3 seconds). Once we select a smaller session group, performance returns to normal.

Our server runs consistently below 30% utilization for CPU, RAM, network, and disk queue stays pretty close to 0 at all times.

On my PC which runs a current gen i5, Chrome generally stays around 10% CPU while actively using Screenconnect, but goes up to a constant 30-40% on all cores when selecting a group of 2000+ sessions.
MannyTC  
#6 Posted : Tuesday, May 2, 2017 4:59:00 PM(UTC)
MannyTC


Rank: Advanced Member

Medals: Bug Buster Level One: Spoon!Level 3: Shirt off your back! Received 25 Thanks!

Joined: 2/19/2015(UTC)
Posts: 256
United States
Location: AZ

Thanks: 8 times
Was thanked: 50 time(s) in 43 post(s)
We only have about 350 access clients and while the performance is not terrible we did see a noticeable slowdown in the initial load of the list that was not there before.
futura  
#7 Posted : Tuesday, May 2, 2017 6:56:42 PM(UTC)
futura


Rank: Advanced Member

Medals: Level 1: Random Act of Kindness! Received One Thanks!

Joined: 1/3/2014(UTC)
Posts: 99
Canada

Thanks: 46 times
Was thanked: 11 time(s) in 10 post(s)
I too have noted that it is running more sluggish as of late... we're on 6.2 with about 300 Access clients.
amihaly  
#8 Posted : Wednesday, May 3, 2017 3:04:19 PM(UTC)
amihaly


Rank: Newbie

Joined: 3/26/2015(UTC)
Posts: 1
United States
Location: Ohio

our company is also experiencing the slowness with new update. I have 754 clients on selfhosted. With testing of this i have found out if i do not upgrade clients to latest performance is much better. version 6.2.12963.6312
thymesoft  
#9 Posted : Sunday, May 7, 2017 11:13:47 PM(UTC)
thymesoft


Rank: Newbie

Joined: 2/12/2016(UTC)
Posts: 6
Australia

Our company has about 250 access sessions, and we've been having slowness in some of them since version 6.1. We are using the cloud hosted version of ScreenConnect, so we have no direct control over the hosting details.

It's always the same sessions that are slow, but those are always VERY slow. Click the mouse, and wait (sometimes for minutes) before seeing the final result.
We were hoping version 6.2 would fix this, but it has not improved. These systems have all been using ScreenConnect since sometime in version 5, at which time they were working ok.
We've not found any pattern to the systems that are affected, but it's always the same systems. They are not old systems, and they have between 4 and 16 cores, with between 4GB and 16GB RAM.

We found a few things by experimentation that may be useful:

1. Down-grading the client version installed on these systems back to v6.0 seems to help quite a bit. The test process was: Install an alternative product to retain some kind of access. Uninstall the ScreenConnect client. Reinstall the old version 6.0 ScreenConnect client. Turn off the alternative product. The ScreenConnect speed then is quite usable. Reverting to v6.1 (or v6.2) after this returns to the extremely slow performance as before.

2. We also found that if we click the mouse, and then disconnect the host, and then immediately reconnect the host again, then the initial screen loads quickly and has caught up and is ready for the next mouse click. This is true even at the slowest times for the slowest responding operations. Before we found out that we could get a reasonable result by downgrading the client to v6.0, our process on these systems had degraded to a continuous stream of click-disconnect-reconnect, just so that we could get anything done. (For an engineering perspective, the impression is as though, while connected there is a queue of screen updates that take a long time to transmit from the client, but when disconnected the queue gets cleared and restarted from fresh when we reconnect. The initial screen load on reconnecting is almost instant.)

3. We've tried some other products (TeamViewer & Splashtop) on those same systems, and those work fine, at a quite respectable speed.

Edited by user Sunday, May 7, 2017 11:14:46 PM(UTC)  | Reason: Not specified

solveit2  
#10 Posted : Tuesday, May 16, 2017 4:17:22 PM(UTC)
solveit2


Rank: Guest

Joined: 5/16/2017(UTC)
Posts: 1
Ireland
Location: Clonmel

Hi
+1 for the very poor performance of 6.2.12963.6312
25% cpu usage on most remote clients.
Ben B  
#11 Posted : Tuesday, May 16, 2017 5:03:28 PM(UTC)
Ben B


Rank: Administration

Medals: Level 2: Lent a Helping Hand! 10 Thanks!

Joined: 10/2/2015(UTC)
Posts: 307

Thanks: 1 times
Was thanked: 63 time(s) in 58 post(s)
Originally Posted by: solveit2 Go to Quoted Post
Hi
+1 for the very poor performance of 6.2.12963.6312
25% cpu usage on most remote clients.


Good afternoon,

There are two issues described in this thread.

The first has to do with host page performance degradation after upgrading to 6.2.12963, especially when working out of larger session groups. This issue has been registered with development and is currently under investigation. More information, including a potential workaround, can be found in this technical bulletin:

https://docs.connectwise..._issues_on_6.2_host_page

The second issue under discussion has to do with client performance. We identified an issue with the way the client CPU resource consumption is throttled in 6.2.12963; guest clients installed on machines running older versions of Windows (i.e., Windows 7/2008 and below) with four or more cores consistently use ~25% CPU resources. Development has implemented a fix for this issue, and it should be included in the next release.

Cheers
ScreenConnect Team
thanks 2 users thanked Ben B for this useful post.
buxtonffemtp on 5/16/2017(UTC), marktoo on 5/16/2017(UTC)
buxtonffemtp  
#12 Posted : Tuesday, May 16, 2017 5:20:10 PM(UTC)
buxtonffemtp


Rank: Guest

Joined: 5/16/2016(UTC)
Posts: 1
Man
United States
Location: Maine

Thanks: 1 times
I am happy to hear that we are not the only ones experiencing issues on this last upgrade. I have been chasing an issue that was severely degrading access to a server with the newest verson installed. I uninstalled it completely off of the server and it was light a light switch in speed and performance. Screenconnect was eating over 30 % of resources on a 2011 SBS Server. Extremely disappointed, but am looking forward to the fix. There does not seem to be any rhyme or reason that it is fine on some but devestating on others.

This is really the first hiccup our company has had since we have had this software clsoe to 6-7 years.
marktoo  
#13 Posted : Tuesday, May 16, 2017 8:48:45 PM(UTC)
marktoo


Rank: Advanced Member

Medals: Level 1: Random Act of Kindness! Received One Thanks!

Joined: 3/29/2015(UTC)
Posts: 99
United States
Location: Metro DC

Thanks: 28 times
Was thanked: 8 time(s) in 7 post(s)
Glad to know these two issues are registered and being worked on! We're still at 6.1... and, based on these informative posts will wait until 6.3 to upgrade our instance. Thanks to everyone for your input!

Mark
technician@uescomp.com  
#14 Posted : Wednesday, May 24, 2017 5:45:17 PM(UTC)
technician@uescomp.com


Rank: Newbie

Joined: 1/18/2016(UTC)
Posts: 1
Location: US

1000 endpoints and extremely slow on all connections. Since that last 2 updates it has gotten so slow it is almost unbearable at times. When both ends have 35x5 connections or faster it should not be this slow. Turning remote session to black and white does not make a difference.

Viewing sessions in the browser is fast. Once you connect to a computer/host it runs at dial-up speed. Click, wait 10-20 seconds to see result, click and wait some more....

Edited by user Wednesday, May 24, 2017 5:48:55 PM(UTC)  | Reason: Not specified

RobertR  
#15 Posted : Friday, December 15, 2017 2:13:20 PM(UTC)
RobertR


Rank: Guest

Joined: 8/22/2017(UTC)
Posts: 2
United States

I know this is an old post, but I am having the same issue where SC is pegged ~ 25% CPU on only some of my endpoints, which generally seem to be Windows Server 2008 and 2008 R2 clients. I'm on the latest SC build, and LogMeIn and RDP work fine, whereas SC is unusable because it is so slow.

Have there been any developments on this issue?

Thanks,
Robert
RobertR  
#16 Posted : Friday, December 15, 2017 2:22:48 PM(UTC)
RobertR


Rank: Guest

Joined: 8/22/2017(UTC)
Posts: 2
United States

OK, never mind. I just saw one of the other posts regarding SC and Matrox drivers. I disabled hardware acceleration and SC works great now. This is why I love support forums!

Thanks,
Robert
Users browsing this topic
Similar Topics
Major Performance Issues - New Client (Bug Reports)
by p1it 5/25/2015 7:15:30 AM(UTC)
Video Related Performance Issues (Bug Reports)
by CMITWexford 11/10/2014 10:44:24 PM(UTC)
Server Performance issues after upgrading from 3.2 to 4.2 (Advanced Customization)
by foodtec 6/10/2014 11:32:30 AM(UTC)
Forum Jump  
You cannot post new topics in this forum.
You cannot reply to topics in this forum.
You cannot delete your posts in this forum.
You cannot edit your posts in this forum.
You cannot create polls in this forum.
You cannot vote in polls in this forum.