How to deal with MEGA's "Temporarily Unavailable" Message
Note: This is still based on a working theory, but I believe it to be correct based on my experiences which have encountered this issue more than once.
Disclosure: I work in IT so I do know what I'm talking about re: managing very large amounts of data; the below describes a real world data management strategy used in many industries....of course I could be wrong since mega doesn't explain their system, but this is what I think until I see evidence to the contrary.
It is a common strategy when managing huge amounts of data to put less used data onto cheaper storage meant for archival use (meaning not immediately available, such as on tape). If something was moved to tape behind the scenes, you would get a message like mega displays about being temporarily unavailable.
I have found the best way to deal with this is to be using jdownloader, which will automatically keep trying to download the file every 5 minutes (or something like that). If you leave jdownloader running, it will eventually get the file, though it migh take up to two days. Just let it keep running, preferably 24/7 until you get the file.
The reason it might take so long is because mega is probably using an algorithm based on number of requests to decide when to bring back the file you are requesting. It also might be putting it into a queue with other requested files, so you want to keep on requesting it to make sure it doesn't "lose its place" in line.
The way mega chooses which files get sent to tape doesn't seem to make a lot of sense, in that it seems only one or two parts of a rar seem to be affected by this sometimes. But this behavior is nonetheless very consistent with the mechanism I've described. What probably causes it is that when you upload your files they are getting placed on many different drives; mega would want to fill every one of their hard drives to full capacity, so it might put small files onto random hard drives that are almost full. This would explain why only certain parts go missing, since the decision to put something into archival storage might be based on the age or requests for files for the entire hard drive, not just the individual file.
A more consistent example is almost any online banking site. Most let you download your past statements, but if they are old, they will tell you it will take XX time before you can download them once click on the link. That's because it has been moved to archival storage (tape) and has to be restored to online storage (hard drive). I think this is the same thing that causes megas temporarily unavailable message, and the best way to deal with it is to keep on requesting it automatically, using jdownloader.
PS - if someone reports to you that some specific file is temporarily unavailable, you might log onto mega and see it reported as online. That doesn't mean it really is online. (Archival systems use file "stubs" that will initially report the file is online even if it is not. thats so they can keep all files visible to users even if it has been moved to tape behind the scenes.) Try to actually download it yourself and you will probably get a temporarily unavailable.
The best way to get it back online would be to set up jdownloader yourself to keep on asking for the file. In the meantime, re upload the part as a different file (if you can), for which you can then give the link out as a temporary measure until the original gets back online.