high cpu usage on gigahub

[Gptechman]

anyone got high cpu usage on their gigahub? since update 2.13 it always been high, this is unusual
cpu load has technically never gotten high since i signed up for fiber, but now it its unexpectedly.
yesterday I've seen the bar go up to 90%

 

[dks]

Thank you for your feedback. No, that's not normal. How many devices do you have connected to your Giga Hub? What is your speed profile? 

[Vanadiel]

Looks like a maxed out gigabit connection at 945 Mbps. Ping times look normal also.

The CPU load is abnormally high though.

I don't think the home hub has a command prompt. If it does, it should be easy to figure out which process or processes are causing the high average CPU load, by using top or glances at the command prompt.

[Gptechman]

?

[Anatharias]

I got a decent amount of devices connected. Though, when it happens, it often happens early in the morning or during the night, but around the 3-4 weeks mark after a restart. At that time, only my Synology NAS and a Linux laptop (plex) are running, along with a decent amount of Wi-Fi devices like thermostat.

I deactivated uPnP... how can we troubleshoot this further ?
also worth mentioning that were Bell customers since they added Fibre to the neighbourhood and this behaviour only happens recently, like 3-4 month ago, and I don,t recall adding or changing anything of relevance to the network.

it becomes so unresponsive that the modem web page is no longer sending any data !

[Vanadiel]

I don't believe it has a command prompt, so it will be difficult to further troubleshoot this issue on your end.

It would be something someone from the Bell firmware team would need to pick up and look at, as they must have the ability to get to a command prompt to see what processes are taking up all this CPU time, and why.

For example, on my Mikrotik router and can use /tool profile cpu=all at the command prompt to see all the processes that are running and what load they pose on the system.

Columns: NAME, CPU, USAGE
NAME CPU USAGE
www 0 0%
console 0 0%
firewall 0 0%
networking 0 2%
management 0 0.5%
telnet 0 0%
bridging 0 0%
unclassified 0 0.5%
cpu0 3%
www 1 0.5%
ethernet 1 0%
firewall 1 0%
networking 1 0%
management 1 0%
unclassified 1 0%
cpu1 0.5%
networking 2 0%
management 2 0%
unclassified 2 0%
cpu2 0%
www 3 0%
console 3 0.5%
networking 3 0.5%
management 3 0.5%
unclassified 3 0.5%
cpu3 2%

[Gptechman]

oh, right i know what you mean by yeah on bell's firmware "maybe" they might have this feature on a tech's end.

but obviously they remote into the modem's daily so probably a remote monitoring for performance only.

[dks]

Thank you for the additional information. This will require forwarding to Bell. Unfortunately we have no connection with ant teams other than staff who monitor the forum, so they will have to pass the information along. 

[WayneYe]

I’ve been experiencing an issue with my internet connection where it becomes extremely slow at times. During these periods, it can take several minutes to load a webpage. The issue usually lasts anywhere from a few minutes to several hours and tends to occur in the evenings.

I’ve tested the connection through the GigaHub console during these slowdowns and noticed that the ping to Google exceeds 2000ms. Additionally, the GigaHub shows a consistent 100% usage rate, which seems very unusual.

Could you please help me investigate and resolve this issue?

[WayneYe]

[dks]

Thank you for your question. A ping of 2000ms suggests the issue is network congestion. When this happens, if you could run a traceroute from the Giga Hub desktop and post the results, it would be helpful in understanding where the issue might be. 

The issue of high Giga Hub CPU usage has been noted by users, including myself, in the past. It has been drawn to Bell’s attention. It does not appear to have an impact on your user experience. If you are having it affect your user experience, those symptoms would be helpful to know, as well. 

Finally, running the Virtual Repair Tool on the support page may also help in diagnosis of any other issues. Thanks for sharing your experience. 

[WayneYe]

Thank you for your earlier response and recommendations. I performed a traceroute from the GigaHub console during the most recent occurrence of the issue, but the result consistently shows: Error: Max hop count exceeded.

Additionally, the issue persists with extremely slow internet speeds, making it difficult to browse or load webpages effectively. The high GigaHub CPU usage also coincides with these slowdowns.

I have run the Virtual Repair Tool as suggested, but the problem has not been resolved. Could you please provide further guidance on how to address this issue?

Thank you for your continued assistance.

[dsmider]

Please inform your technicians to please fix the ongoing problems associated with firmware 2.13 on the bell Gighub 4000. Ever since this update was pushed its been problems with persistently high CPU usage on the router for no reason, random issues with high latency and packet jitter and Advanced DMZ causing packet loss and not working correctly. Its a firmware problem, why is this not being addressed ?. Please have it fixed or at least allow users to use their own SFP+ (XGS-PON/GPON) modem.

[dks]

Thank you for your feedback. The concerns you have raised have been noted by other users. A forum search should bring you other users experience and Bell's response. 

For technical reasons the ONT was brought into the Giga Hub. It is doubtful that will change as it allows for a user installation option. You might also engage with other users on Reddit about your technical issues.   

[dsmider]

I've run a continuous ping to the closest node I can think of to rule out congestion on the network. A ping to Bells local IMS front end that handles Homephone(VoIP/SRTP) and IPTV is only a single node away and shows the same issues as any other network path. The path to this server has elevated or maximum QoS prioity enforcement Its either a modem issue(likely) or a media fault of some kind but no frame errors are recorded on the OPTICAL0(SFP) interface so I don't think that's the problem. I've observed that the problem does seem to occur more frequently when the CPU usage starts to hit 90-100% and generally sits at around 50% - 65 %, never anything less than that.
.

[Raka]

Also experiencing high cpu usage. Ocassional spikes to 100%.

Same number of devices and connections no changes. What can be done to lower the usage? 

I set transmit power of 2.4ghz and 5ghz to 50% does that help? I checked and monitored and saw I believe a 5% drop so maybe not an outlier. But I still managed to see a 100% spike.