We currently watch the size of our relay cache folder and increase the size accordingly with the setting “_BESGather_Download_CacheLimitMB”. However, if for some reason, the limit is reached and the old cached downloads are removed to make way for the new downloads, we would be losing either our custom software distribution cached installs, or important Linux/Unix downloads that we had to manually download and cache.
What does everyone else do to manage their cached downloads? Is there some other folder we can place the downloads that it can use as a secondary cache repository that it won’t delete?
You wouldn’t “lose” them per se. The relay would simply re-download the cached file from the main server, bumping out the next oldest content on the relay in the process.
I’m not sure what happens if the main server reaches its limit, to your point. I have the size set rather large on the main server so I hopefully never have to find out. Does anyone know what happens then?
Content you have uploaded to the server using the Software Distribution wizard or other console based mechanisms will not be removed, nor will content you have placed manually in the wwwroot folder. Only items that are in the sha1 cache folder are subject to the caching rules. Does that alieveate your concern?
answers, but does not alieveate. My concern is with the manual caching of our Unix/Linux patches. We are currently not using the wizards because the Unix operators don’t have access to the main bes server within the console. So, they manually download the update, rename it to the sha1 value, then I put it in the sha1 folder of the main bes. Their thoughts are, what if the sha1 folder “runs over?”. Will their downloads be removed? Is there a second location that they could be stored that would be safe. From what I am hearing, they will be removed since we are not using the wizards to cache them.
Yes, they will be removed from there eventually based on the caching rules. If you want the downloads to persist, you can place them anywhere inside the wwwrootbes folder and then craft your download actions to reference the root server. A good way to do that, especially if you are using a DSA server, is to just use the loopback address. So the download command would look like: