On Wed, 25 Oct 2023 06:23:25 +0000, Timothy Sipples <sipp...@sg.ibm.com> wrote:
>Jon Perryman wrote: >> Since there are lots of reasons, can you name 3 beyond those I mentioned? >> (backup to cloud) >How about everything else works this way (including z/OS), Googles cloud backup/recovery is very different from IBM z/OS. The TS7700 is a private cloud and z/OS believes the TS7700 is a tape drive which does not use any cloud API nor TS7700 cloud object. We know that Google starts a Linux image when you request recovery which is not cloud object API unless Google is deviating from the spirit of cloud object. Maybe multiple cloud API's are involved but that is internal to Google. I'm sorry if you think I'm insulting you or IBM. To the contrary, IBM builds exceptional solutions. There are lots of exceptions because there is more than one way to solve a problem. The first few design drafts of "the cloud" were met with great excitement because of IBM participation that made "the cloud" sound like an ad for z/OS sysplex. Many of those features were dropped after other vendors realized it would take many years to implement those features. > they don't want z/VM to be different/exceptional (not in this respect), > and they have greater confidence/assurance that their backups will > be better secured/encrypted and better protected from local disasters this > way? No IBM z system has cloud backup. You can't backup z/OS to any other cloud than that provided by TS7700. TheTS7700 completely hides cloud backup concepts. >Why does Iron Mountain exist and thrive? It�s the same basic set of reasons. It exists because IBM does smart design instead of convenient design. Take for example z/OS datasets where a cobol programmer easily uses various access methods with a simple changes to the FD, READ and WRITE. On the other hand, Unix requires the programmer to rewrite their programs using NOSQL and SQL API's to achieve the same results. >> I suspect Ayre is saying cloud but I doubt Ayre has a specific cloud solution >>in mind nor implied "cloud object storage". > Cloud object storage is what the public commercial clouds (also) all provide > for backup data storage/retrieval. Cloud object storage is the service, > and then that service can be provided by public commercial clouds (e.g. > Amazon S3), > privately hosted cloud object stores, or some combination. Not to downplay the importance of cloud objects but what makes it the only solution he should consider? I agree this is an obvious solution.but my point is that "cloud" is not a requirement but considered his obvious solution. Until proven otherwise, we have to assume everyone involved has a birds-eye perspective without specific knowledge of things like cloud object. >>Implementing a new feature request takes time (Potentially years). > Potentially, but that�s not a reason to skip filing a feature enhancement > request. > It�s a great reason to file a request now rather than later. If this a feature that Ayre would implement, then asking doesn't hurt other than wasting valuable resources that IBM continues to reduce. I suspect that z/VM backup has a similar design to z/OS backup. Ask yourself why IBM didn't implement cloud object for z/OS DFDSS instead of building a TS7700 with cloud objects. Imagine all your files are migrated and every "open" causes a recall. Recall then causes the restore (a cloud object). I'm not familiar with cloud object API but this could cause a lot of network processing and processing that was previously handled by controllers. Worse yet, you potentially have huge delays that IBM avoids in their hardware design. >>The obvious problem is maintaining a TS7700 in another country and >>moving it if that country becomes a problem. > No more or less obvious than the already extant requirement to maintain a > suitably configured IBM Z server > with sufficient storage in an alternate site to restore the data, recover, > and resume service. "requirement to maintain" is part of what I'm talking about. The company must maintain IBM equipment where they don't have trained staff. All other cloud services (e.g. Google, AWS, Microsoft, ...) have trained staff. Many of these clouds encompass the world. You move data from one location to another without hardware considerations. Better yet, they will most likely have staff that speaks your language. All your points are valid. Each company has requirements. In the case of Ayre, the company doesn't want to maintain anything. It sounds like they want a disaster recovery solution that is as simple as offsite tape backups. ---------------------------------------------------------------------- For IBM-MAIN subscribe / signoff / archive access instructions, send email to lists...@listserv.ua.edu with the message: INFO IBM-MAIN
