Update and Relay servers

Hello David Peterson,

Please have a look at this article to ensure all the pre-requisites were met

More details here 

Regards,

A few devices are showing on Sophos Central as using the local server, 4 of them. Another 78 are using a particular remote site, there are not nearly that many systems there on Sophos. Before, they were using a different server, which alphabetically came before the one they're using now. We disabled the update cache on that server and they ended up moving down the line to another remote server.

Since there are devices on the local cache server I'd think it'd meet the requirements, and looking at them I don't see any it doesn't meet.

A few devices are showing on Sophos Central as using the local server, 4 of them. Another 78 are using a particular remote site, there are not nearly that many systems there on Sophos. Before, they were using a different server, which alphabetically came before the one they're using now. We disabled the update cache on that server and they ended up moving down the line to another remote server.

Since there are devices on the local cache server I'd think it'd meet the requirements, and looking at them I don't see any it doesn't meet.

Hello David Peterson,

I would recommend to follow the first article I provided. Aside from the requirements, it includes troubleshooting steps and suggestions on how to check what's updating from where.
Are the machines able to access the local server you are referring to? Are there any restrictions in place? Also, is the update cache healthy and running? (The end of the article covers this as well). 

Please see these sections (first) to try to pinpoint what may be affecting your environment:
Can I check if any devices are updating from an Update Cache or communicating via a Message Relay? 

How do devices update from a cache? 

How do devices determine which Message Relay to use? 

Let us know how it goes please.

Regards,

I think I figured it out, we were thinking the calculation was on network speed and latency, not just IP number. So a remote site with a 3mbps connection has an IP range closer than our server range at the main office. So, this would work at remote sites, if we have a server at each one (assuming it's not down). That won't work at the main site, and we can't be manually assigning an update cache per machine.

How can we ensure that systems only use a local cache?

The approach I used was simply to block TCP port 8191 on the firewalls between our remote sites.  This means that a machine at one site is only able to see the local update cache server, and if it moves to another, it can only see the one there.

Yeah, just seems to be a bit of a wonky fix to get around bad logic in the product to be honest.

Hello David Peterson,

If the article options do not help, you can file a case with support for further assistance.

If you would like to request a new feature / submit an idea, then please visit this site.

Regards,