ERROR: Could not find a source for updated packages [0x00000071]

Hello Mantas Lenza,

the Secondary is tried when the Primary can't be contacted, thus UNC as fallback for HTTP is rare. The message means that both failed - for whatever reason. If you don't have updating errors mentioning the HTTP location this would mean that always both fail. Does it happen only for some computers? Laptops perhaps? Is it really random and how often do you see these errors?
BTW - why 180 minutes instead of the default 10? With a short(er) interval these errors wouldn't have much impact (and the alert is cleared on a successful update).

Christian

Thanks for answer, I will try to answer to your questions.

I'm using primary update server HTTP because my organization uses Linux OS (Amount of PC's with Linux OS increasing very fast) and UNC doesn't work for Linux, that's why I'm using Secondary update server for Windows.

Maybe for some computers this happens often than to other computers (I always look at updating history of these computers). No, there are not only laptops.

At first I'm using update 180 minutes because of network traffic and second that short updating time dramatically increases updating errors. Now I set these to 45 minutes and looking how it works)

I have couple questions:
1. When Sophos downloading updates they first check and only if update needed then downloading?
2. Could you help or explain why my Windows Computers can't set scheduled scan tasks?  I have tried troubleshoot using community.sophos.com/.../113070 but stuck on this step  "Check if the schedule task has been created on the client" (not works with this /s 127.0.0.1, without this parameter everything is good task created succesful.

Hello Mantas Lenza,

Linux OS
should be able to update via UNC as well, though depending on the distribution it might be necessary to install the cifs-utils or equivalent.

network traffic
when checking for updates the endpoint uses catalogs (the .upd and if required .dat files) to assess whether something and what has changed. So most checks should result in only minimal traffic. Traffic volume is normally not a concern, it's establishing and taking down connections (and this has mostly an impact on the server). As you can expect actual updates several times a day an interval of 180 minutes will only add to updating latency (i.e. increase the time until detection data updates are available on the endpoint). 
short updating time dramatically increases updating errors
As mentioned the server might be the bottle-neck - but for meaningful statement knowiledge of the nature of these errors would be required. 

Please start a new thread for the scheduling issue.

Christian

Hello Mantas Lenza,

Linux OS
should be able to update via UNC as well, though depending on the distribution it might be necessary to install the cifs-utils or equivalent.

network traffic
when checking for updates the endpoint uses catalogs (the .upd and if required .dat files) to assess whether something and what has changed. So most checks should result in only minimal traffic. Traffic volume is normally not a concern, it's establishing and taking down connections (and this has mostly an impact on the server). As you can expect actual updates several times a day an interval of 180 minutes will only add to updating latency (i.e. increase the time until detection data updates are available on the endpoint). 
short updating time dramatically increases updating errors
As mentioned the server might be the bottle-neck - but for meaningful statement knowiledge of the nature of these errors would be required. 

Please start a new thread for the scheduling issue.

Christian