I have a roughly 30 GB archive split into 15 parts, with no meaningful compression. Extracting it normally requires another 30 GB of free storage, which I don't currently have. Is there any way to unpack the archive gradually and delete each part as it is processed, so the archive doesn't occupy the full amount of space during extraction? This could become a serious issue with even larger archives.
5 Answers
Some tools can pause between archive volumes, which might let you process volumes one at a time, but deleting a volume before extraction has fully finished can corrupt the operation. If anything fails, you may need to obtain the missing volume again. Extra storage is the reliable answer.
If you have another computer, a tar archive can sometimes be streamed over SSH and extracted directly on the other machine, without first creating a complete temporary copy. A RAM-backed temporary filesystem is another possibility if you have enough memory, but neither option is useful for a 30 GB archive on typical hardware.
The practical solution is to extract to another drive, such as an external HDD, and then delete the archive before moving the files back. A temporary drive or borrowed storage would be much safer than trying to modify the archive during extraction.
Generally, no. Archive tools need the archive data to remain intact while extracting, and deleting files from an archive usually means creating a new copy first. That requires the original archive and the replacement archive to coexist, so it doesn't solve the storage problem.
There are a few special cases, but they don't really fit this situation. Individual files can sometimes be extracted and then removed from an archive, provided they aren't part of a multi-volume or solid archive. Removing them may still require rebuilding the archive, though, and split archives make the process especially risky.
That makes sense. I was hoping the archive could be consumed piece by piece, but it sounds like that isn't safe or generally supported.