Timeline for Is it a bad practice to store large files (10 MB) in a database?
Current License: CC BY-SA 3.0
14 events
| when toggle format | what | by | license | comment | |
|---|---|---|---|---|---|
| Jun 9, 2020 at 14:54 | comment | added | Thomas | @Wecherowski - What you seek is a file versioning system. Even a source control system like git can be used for file versioning. OpenVMS for example has a long history but there are third-party tools than you can get that will do this for you too. It depends greatly on the functionality needed. | |
| Jun 7, 2020 at 18:33 | comment | added | Wecherowski | @Thomas haha, I somewhat expected it to be trickier by the fact that people never mention file server backups in this discussion :-) Can you give me some references or search phrases to dig into this topic? I have close to no devOps experience, searching stuff like "file server database synchronized backup" didn't yield any useful results for me ... (I'm not even sure, whether it has to be synchronized or whether there're other ways ...) | |
| Jun 7, 2020 at 16:48 | comment | added | Thomas | @Wecherowski - This is one of the gotchas with having the files outside the db: you have to devise your own backup strategy for the files. There are plenty of solutions. S3 for example has native versioning that can be enabled so that you don't really have to backup the files. Anytime the file is changed, it is actually versioned under the hood. Of course, accessing those backups and whether those older versions are visible to the user and truly deleting files are all new complications that you have to handle. The "best" answer can depend greatly. | |
| Jun 4, 2020 at 23:04 | comment | added | Wecherowski | When storing files in a DB your backup strategy is fairly simple - that's great. But how would I backup my file server and database in a synchronized fashion? | |
| Oct 26, 2016 at 16:08 | comment | added | B Seven | @BillyONeal - In a way, your comment was the best answer. You can have all the benefits of DB storage, but none of the problems. | |
| Jun 1, 2012 at 13:20 | vote | accept | B Seven | ||
| May 31, 2012 at 16:20 | history | edited | Thomas | CC BY-SA 3.0 | Spelling mistake. Clarified that I'm referring to cloud storage and not a CDN. |
| May 30, 2012 at 21:17 | comment | added | Thomas | @BillyONeal - However, I think this is bad choice of words on my part and I've adjusted my answer. Specifically, if you want to use cloud storage (and then perhaps use a CDN with your cloud storage), you can't do it natively with the database storage solution. You would have to write a synchronization routine to pull the files from the database and then send them to your cloud storage provider. | |
| May 30, 2012 at 21:16 | history | edited | Thomas | CC BY-SA 3.0 | added 8 characters in body |
| May 30, 2012 at 20:22 | comment | added | Billy ONeal | Erm, the whole point of a CDN is duplication. CDNs merely cache the target of a web address -- the only requirement is that there's an HTTP host serving the content, and that the content changes rarely. (How on earth is the CDN supposed to tell where you pulled the image from anyway?) | |
| May 30, 2012 at 20:09 | comment | added | Thomas | @BillyONeal - You can't use a CDN and store the file in the database. Unless you are OK with duplication, you can't have both. | |
| May 30, 2012 at 17:55 | comment | added | Billy ONeal | Why can't you use CDNs? This is a supported scenario with pretty much every CDN I've ever heard of. | |
| May 30, 2012 at 5:41 | history | edited | Thomas | CC BY-SA 3.0 | added 713 characters in body |
| May 30, 2012 at 5:32 | history | answered | Thomas | CC BY-SA 3.0 |