Open Fantu opened 2 years ago
Hi, @Fantu !
I haven't seen any public responses to the latest version, I suspect that someone who was interested missed it due to some patches with incorrect titles.
Yep. Indeed, there is no reaction. I think that everyone who wanted to see the latest patchset did it, regardless of whether he is the recipient of the letter or not.
I am currently preparing an updated standalone version of the module for one of the next releases of our product. I want to move a some ideas from the upstream version to it. This will improve performance.
Even if the blksnap module is never accepted into the kernel, the work has not been done in vain. The code of the blksnap module only gets better from this.
I will definitely return to work on the upstream version later. In working on the next patch for upstream, I think I will focus on more detailed testing and performance measurement.
Hi, this patchset is only for testing on 6.9 but not for submit, right? A v8 will be prepared for 6.10 to submit?
Hi, I don't know yet when I will write e-mails for maintainers. I would like to test more.
Hi, any update on this Topic? What is missing to get the blksnap module into the linux kernel? Regards Flo
Hi.
What is missing to get the blksnap module into the linux kernel?
And that's a good question.
As usual, the most valuable resource is missing - time. But this is a matter of priorities. Well, the most important of them is: "Why?". I know what Veeam needs it for, but it's completely unclear to me if the upstream needs it. I have formed the opinion that this is not interesting to the Linux community. Other backup vendors are not involved in the patchset discussion. Everyone seems to be quite satisfied with dm-snap.
Nevertheless, I plan to make at least one more attempt. It is possible that this will happen this autumn.
Hi.
What is missing to get the blksnap module into the linux kernel?
And that's a good question.
As usual, the most valuable resource is missing - time. But this is a matter of priorities. Well, the most important of them is: "Why?". I know what Veeam needs it for, but it's completely unclear to me if the upstream needs it. I have formed the opinion that this is not interesting to the Linux community. Other backup vendors are not involved in the patchset discussion. Everyone seems to be quite satisfied with dm-snap.
Nevertheless, I plan to make at least one more attempt. It is possible that this will happen this autumn.
Thanks for your honest answer! regards
I would like to contribute more, but I lack the time and I lack sufficient knowledge for the parts of the kernel involved (and the time to learn them). Regarding the interest I think there is more people than think (I mean mainly users), but it is difficult for them to find blksnap and those who find it do not know how it could be useful, and therefore you find mostly messages from users asking for help for new kernel version support (for external module), trying the patches for upstream without enough knowledge to use them, or asking for the status for upstream (there are also in other forums, blogs etc...). Regarding other backup solutions, they mainly concern commercial software and companies in most cases only look at their own interest, calculate very narrowly, stay with something of poor quality rather than contribute even a little to something that also benefits others. Basically, this reasoning (which I consider stupid) could be summarized as: it is better for everyone to stay with worse rather than have better if others also benefit from it. It seems to me that there are still few companies that understand that having more open and collaborative projects can lead to greater quality and potentially also with lower costs, but there are some that are making progress, for example Microsoft has done much more than I would have ever imagined in the last 10 years (compared to the previous opposite thought they had). Regarding the developers involved in the discussions, it seems to me that most of them have a limited vision, calculating only or almost their own uses and skills/knowledge and underestimating the segment of users to whom it would be useful, both companies that need a single, centralized and easily manageable backup system for several devices and smaller situations that also need more complete and simple backups. There is something with other solutions, for example some that are slowly being integrated into various distros (but they are mainly things that require more work and limitations to use them massively), but even there they are going slowly and therefore I think blksnap could still have a significant slice of use, even if not as large as it would have been 5-10 years ago.
Yes, that's right. I agree with every word @Fantu.
Plus, I and the company I work for have pretty little experience working with open source. Perhaps , this is our point of growth. I suppose that more efforts should be made in this direction. This is just my personal opinion.
@SergeiShtepa Hi, thanks for your works on blksnap kernel module, I think will be very useful for improve backup on linux. I'm curious about the status of its possible upstream integration, from https://github.com/veeam/blksnap/commit/85bad8f010d08fb0d08fc2c33e5ca3630dfe2a82 seems you prepared a patches serie for upstream (for send a mail with it) but I not found it posted searching on lkml.org and google. I saw also other series of months ago on other branches but I not found them on upstream mailing list and found only older your works of 1 year ago or more. Thanks for any reply and sorry for my bad english.