4 ms·
Having done startups at 19, 39, and 45, I can say that the sh$t I pulled at 19 is a no go at 45 with kids, but I'm also a lot wiser now. That week of near death
by dgacmu 4y ago
Having done startups at 19, 39, and 45, I can say that the sh$t I pulled at 19 is a no go at 45 with kids, but I'm also a lot wiser now. That week of near death panic due to having had my backups fail as. 19 year old will never happen to me again (knock on wood, backblaze don't fail me now). Etc. You trade some hours for some efficiency.
(There's also the financial risk, which for me is easier to absorb now. At 19 I didn't care, at 30 something I did, and now I'm lucky enough to not have to care much).
- bombcar 4y agoIf you’re worried about backblaze failing you, grab some local storage and do some manual physical offline backups you can take offsite. Even monthlies have a value in the worst case scenario.
- ethbr0 4y agoYour comment made me wonder if there's anyone out there with regularly-taken long-persistence (tape, optical media, etc.) backups in their personal basement. It seems a very specific Venn intersection to utilize technology to a high degree, but also to imagine and prepare for its worst case failure. Maybe the NZ tech-prepper crowd?
- bombcar 4y agoThere are dozens of us! I make regular yearlies that go "basement" and never get overwritten. Has come in useful very rarely, maybe it was worth the cost?
- james_a_craig 4y agoI keep an LTO tape robot around for just exactly that reason. I trust backblaze, but I don't trust them to the point of not writing off a weekly/monthly/yearly tape set.
- bigtunacan 4y agoA company I previously worked at (self-hosted) all production servers were RAID 5, plus continual cloud backups, with nightly backups that rotated weekly and monthly backups that were shipped out of state and stored in a disaster recovery facility in the event of a total catastrophe (earthquake, hurricane, etc). The off-site backup storage was required by several of our largest customers due to the critical nature of our software.
- dgacmu 4y agoI'm not really. We have onsite backups + off-site to backblaze. I'm comfortable with that for where we are. (We run on-prem for most things that involve a lot of data or GPUs.)
- nextaccountic 4y agoIf you worry about backblaze, the solution isn't to have yet another backup, it's to test your backup recovery process regularly The risk from not being able to recover your data because of something wrong with your process (like, you are not backing up all the data you need) dwarfs the risk from backblaze to lose your data
- bombcar 4y agoThat's another type of risk, but if all your backups are in backblaze, you're exposed to "single provider" risk which can become very real very quickly (and not only by backblaze going bankrupt overnight). Of course, that risk may be entirely acceptable; it's all about risk mitigation and trade-offs, not perfect security.
- vram22 4y ago>If you worry about backblaze, the solution isn't to have yet another backup, it's to test your backup recovery process regularly The solution is to do both, or all three (or n), where #3 is another backup process. And to test the backups of all processes by trying to read or restore them (onto empty disks) regularly. I.e., somewhere on a continuum between zero or minimal and the above, depending on how critical your business is. Because Murphy's Law. Edited to add: And how critical your job is (to you). Of course there is much more to the whole issue, like funds, PHBs, etc. - Bildert.