<div style="font-size: 12pt;">Hello!</div>
<div style="font-size: 12pt;"> </div>
<div style="font-size: 12pt;">Addendum.<br>I'll add obervations, which may be related.<br>1) Firefox stucks - no data comes in, network 0b/s.<br>2) Viewing video, video stucks shortly after start.<br>If this just happens, just change the current workspace</div>
<div style="font-size: 12pt;">on the client side (where the windows are) and switch</div>
<div style="font-size: 12pt;">back immidiately: Problem "solved" in the same moment.</div>
<div style="font-size: 12pt;">The one window has its data, the other play the video.<br>Looks like, it has something to do with window manager</div>
<div style="font-size: 12pt;">or something, which is not knowing, where the rigth</div>
<div style="font-size: 12pt;">window is?<br>(I am using published applications most of the time).<br><br>Update.<br>Due to updates, I had to shutdown all and restart.<br>This make verything shine golden and remind me,<br>that immidiately after starting the x2go client (and</div>
<div style="font-size: 12pt;">server!), perf is really well (I described my hw/sw</div>
<div style="font-size: 12pt;">already).<br>Its usually over time, that perf degrades that much.<br>Over time is someting really over days, because<br>I usually do no shutdowns, but hibernate my workstation.<br>What I saw before the restart, was additionally a laggy<br>scrolling behavior in Firefox and it reminds me, that<br>I saw this bevore - looks probably related to the memory<br>leak of the x2goagent. After the restart, scrolling<br>works fine.<br><br>Related.<br>I have two lxc-containers running on my gateway, one<br>for dns, one for a squid proxy. Both works well,<br>but sometimes, perf is in question.<br>I usually run a terminal to the gateway host and<br>monitor this containers using lxc-top. The related<br>thing is: relative high cpu (20%) and growing mem-use,</div>
<div style="font-size: 12pt;">the reason, they must finally rebooted each two weeks.<br>Now, I've disabled x2goserver in both containers and</div>
<div style="font-size: 12pt;">I am running the third day now: 0%(!!) cpu, NO memory<br>leak. They all run debian buster (fully upgraded)<br>on kernel 5.10, x2goserver 4.1.0.3-4.<br><br>May be this helps.<br>So the memory-leak seems to be obvious. In the current,<br>new cycle for my main-use, mem climbed already up</div>
<div style="font-size: 12pt;">from 0.6% to 5.0%.</div>
<div style="font-size: 12pt;"> </div>
<div style="font-size: 12pt;">Regards,</div>
<div style="font-size: 12pt;">Manfred<br><br><br><strong>----- Original Message -----</strong><br> <strong>From:</strong> webman@manfbraun.de [<a href="mailto:webman@manfbraun.de" target='_blank'>mailto:webman@manfbraun.de]</a><br> <strong>To:</strong> <x2go-user@lists.x2go.org><br> <strong>Sent:</strong> Tue, 11 Apr 2023 02:27:46 +0200<br> <strong>Subject:</strong> [X2Go-User] Performance problems with x2go - hickups, pause, probably memory-leak<br><br></div>
<div style="font-size: 12pt;"> Hello!</div>
<div style="font-size: 12pt;"> </div>
<div style="font-size: 12pt;">I probably detected strong problems with x2go, not very easy to explain.</div>
<div style="font-size: 12pt;">I am running a debian bullseye desktop, on it a kvm guest with debian</div>
<div style="font-size: 12pt;">bullseye, both have kernel6. The kvm guest x2goserver: 4.1.0.3-5, the</div>
<div style="font-size: 12pt;">host (my workstation) x2goclient: 4.1.2.2-2. Usually everything updated.</div>
<div style="font-size: 12pt;"> </div>
<div style="font-size: 12pt;">The problem is the performance of firefox, but not generally. It stucks</div>
<div style="font-size: 12pt;">sometimes, up to several minutes and is completely unresponsive,</div>
<div style="font-size: 12pt;">especially(!), if it comes to youtube videos, but most time, it is not</div>
<div style="font-size: 12pt;">video.</div>
<div style="font-size: 12pt;">Astoundigly, in the many cases, it recovers. I am run dstat permanently</div>
<div style="font-size: 12pt;">(in the guest) and usually see a memory bottleneck (mem < 1 GB), not</div>
<div style="font-size: 12pt;">too strong and additionally high cpu usage, load: 6/4/4, but not too long.</div>
<div style="font-size: 12pt;">Often, this box crashes, if mem goes below 500 MB. Known, with firefox.</div>
<div style="font-size: 12pt;"> </div>
<div style="font-size: 12pt;">For unknown reasons, this situation is not stabil (I am using up to</div>
<div style="font-size: 12pt;">500 Tabs), one will never know, when this may happen and pre-actively</div>
<div style="font-size: 12pt;">saveing the session and continue with, say, 50 Tab, doesnt change the</div>
<div style="font-size: 12pt;">problem!</div>
<div style="font-size: 12pt;"> </div>
<div style="font-size: 12pt;">The VM got 8 cores(!) on a AMD EPYC-Rome Processor with max 3.4 Ghz</div>
<div style="font-size: 12pt;">and 24 GB(!) mem, it got two network interfaces, one, to communicate with</div>
<div style="font-size: 12pt;">my host over a local bridge (10-36 Gb/s!), this is the "x2go-channel" and</div>
<div style="font-size: 12pt;">a lan-IF (1 Gb/s, hardware-passthrough) to the gateway (has 500 Mb/s) and</div>
<div style="font-size: 12pt;">nothing overloaded ever. Additionally some diskwrite, but with 600MB/s</div>
<div style="font-size: 12pt;">(in the VM!) this should not indicate a bottleneck.</div>
<div style="font-size: 12pt;"> </div>
<div style="font-size: 12pt;">But since days, while another hickup, I saw, that the fox needs about</div>
<div style="font-size: 12pt;">only 20% cpu, BUT 'x2goagent' neds 100% !! It looks like the hickup</div>
<div style="font-size: 12pt;">appears exactly at those times. I setup a little processmonitor for the</div>
<div style="font-size: 12pt;">x2goagent and this seems to show a memory leak. I append this at the</div>
<div style="font-size: 12pt;">end. I installed 'glances' and that show me the x2goclient as the second</div>
<div style="font-size: 12pt;">mem consumer, growing.</div>
<div style="font-size: 12pt;">What language is the x2goagent written? Looks like a failing</div>
<div style="font-size: 12pt;">heap-management, garbage collection, which needs a lot of cpu?</div>
<div style="font-size: 12pt;"> </div>
<div style="font-size: 12pt;">Some notes would be really welcome!</div>
<div style="font-size: 12pt;"> </div>
<div style="font-size: 12pt;">Thanks,</div>
<div style="font-size: 12pt;">Manfred</div>
<div style="font-size: 12pt;"> </div>
<div style="font-size: 12pt;">--- x2go-"monitor" -----------------------------------</div>
<div style="font-size: 12pt;">(to keep this short, it has been cutted in the middle)</div>
<div style="font-size: 12pt;"> </div>
<div style="font-size: 12pt;">$ while true; do ts="$(isodate -dt)"; xp="$(ps --no-headers -o user,pid,rss,%mem,comm -p 9430)"; echo "${ts} ${xp}"; sleep 3600; done<br>2023-04-05--22-21-29 ncu1 9430 1479916 6.0 x2goagent<br>2023-04-05--23-21-30 ncu1 9430 1491476 6.0 x2goagent<br>2023-04-06--00-21-30 ncu1 9430 1507612 6.1 x2goagent<br>2023-04-06--01-21-30 ncu1 9430 1514504 6.1 x2goagent<br>2023-04-06--02-21-30 ncu1 9430 1545784 6.2 x2goagent<br>2023-04-06--03-21-30 ncu1 9430 1560096 6.3 x2goagent<br>2023-04-06--04-21-30 ncu1 9430 1664844 6.7 x2goagent<br>2023-04-06--05-21-30 ncu1 9430 1684888 6.8 x2goagent<br>2023-04-06--14-06-28 ncu1 9430 1684888 6.8 x2goagent</div>
<div style="font-size: 12pt;">.......</div>
<div style="font-size: 12pt;">2023-04-10--16-06-31 ncu1 9430 2263236 9.1 x2goagent<br>2023-04-10--17-06-31 ncu1 9430 2274848 9.2 x2goagent<br>2023-04-10--18-06-31 ncu1 9430 2305704 9.3 x2goagent<br>2023-04-10--19-06-31 ncu1 9430 2352964 9.5 x2goagent<br>2023-04-10--20-06-32 ncu1 9430 2352964 9.5 x2goagent<br>2023-04-10--21-06-32 ncu1 9430 2406988 9.7 x2goagent<br>2023-04-10--22-06-32 ncu1 9430 2446112 9.9 x2goagent</div>
<div style="font-size: 12pt;">======================================================</div>
<div style="font-size: 12pt;"> </div>
<div style="font-size: 12pt;"> </div>
<div style="font-size: 12pt;"> </div>