It's a bit frustrating to spend my days concentrating on writing for the beta of the next version of DPM, yet not being able to say anything about it here. It makes perfect sense, of course -- not mentioning features or functions prematurely. After all, if I wrote that we were thinking of having V2 generate an alert whenever a file is deleted from a server and a zillion companies cancelled their plans to order V2 because they (rightfully) consider that a terrible plan...or that V2 would perform on-the-fly defrag of each protected server, and somebody made their business plans counting on the availability of that feature...then V2 didn't generate an alert and didn't defrag server volumes -- well, you see the problem. So, I can't talk about V2 yet. (Just for the record, V2 is not generating an alert for each deleted file or performing on-the-fly defrag while backing up data...I made those up.)
So, let's talk about prototyping instead, just because the User Interface Engineering blog posted an excellent collection of resources yesterday. I wrote a bit about our usability testing last year (has it been that long already?). We used a prototype then, although it was a bit more sophisticated than the paper prototypes mentioned in this article. Although there are great arguments for paper prototyping based on costs and time and efficiency, I think the emotional one is key. As the author explains:
"Modifying a paper prototype is much less painful than for the developers than modifying an actual product. With a real product, because of the substantial amount of work they've put in, the team has an emotional investment in the status quo and will naturally tend to "defend" their design. Even when the team clearly understands the need for changes, it's tough to throw away all that hard work. In contrast, because paper prototypes are so easy to create and modify, there is less invested effort to defend. As a result, development teams become more flexible and willing to try new ideas."
When I read Jeff Alexander's announcement of a new TechNet search engine, my mouse practically leaped to click the link - a better way to search TechNet is something we've all wished for.
I went there. I tried it. I was impressed. The default is the TechNet tab, and it gave me great relevant results. Then I clicked the KnowledgeBase tab and that was full of relevant links. Then I clicked the TechNet Blogs tab and went whoa!
I'm constantly using search at work to locate information on TechNet on other products that I might need to reference when writing about DPM, and this is going to be such a time-saver. Try it!
A recent email from an escalation engineer in England (excuse the alliteration) described a scenario that I think is a great example for understanding data recovery and maximum data loss in DPM. With his permission, let me quote the scenario for you:
Assume an important data file on a protected server that is changing several times an hour.Assume standard synchronisation schedule of once an hour (assume on the hour).Assume standard shadow copy schedule of 3 times a day (08:00, 12:00, 18:00).Scenario 1:
At 13:05 the important data gets corrupted and needs to be recovered.
At that point, the latest data recovery copy available is the 12:00 shadow copy
OR the 13:00 replica. However, to restore from the replica, the administrator would need to force an immediate shadow copy and use that. Administrators who don't fully understand this risk missing being aware of the 13:00 copy and are likely to restore the less up-to-date 12:00 version.
Scenario 2:At 13:55 the important data gets corrupted,
but this isn't noticed till 14:05.In this case the latest 14:00 replica is 'bad' as it is a replica of the corrupted data. Hence in this case, the administrator MUST use the 12:00 shadow copy and not the 14:00 replica (or any 14:05 forced shadow copy).
If the forced 14:05 shadow copy is mistakenly restored, then of course as soon as it is noticed that this is also corrupt they can restore from the 12:00 shadow copy. However, their backup is approximately 2 hours out of date compared to what they possibly expected when the system quotes "Maximum data loss: 1 hour".
The moral of this is that it is important that the administrator understands these concepts and is able to determine which is actually the latest valid backup copy of their data they have, and how old it is, whenever they need to do a restore.
I'm on the new version of Outlook Web Access (OWA) now, which I've been looking forward to ever since I read about some of the new features in Beta 1 on You Had Me At Ehlo (in my opinion, one of the best blog names ever). There were no burps or stumbles in the move over, at least not on my end. But there are a few things I'll have to get used to. One of them is the grouping of emails in the Inbox -- can't seem to get rid of it like I can in Outlook. But what I think is a true oddity is the date in the email list. Take a look at a sample from my Inbox:
I've sort of figured out the rules.
What I haven't figured out, not even sort of, is the thinking behind all of that!
If you've used DPM at all, you're familiar with the process of starting protection of data:
What has happened for several of our customers is that they've tried adding new members to the group or making changes to the protection group settings before step 2 above is complete. So just a reminder - until replica creation is complete, you can't make changes to the protection group. As soon as replica creation is complete, you can add members and change settings all you want!
I can't remember the last time I was this excited by a software program.
I will go to all sorts of lengths to avoid opening a .pdf file -- the experience is just too slow, and it basically renders my computer useless while the opening is going on. Then there's the tortoise-pace of navigating through it. No, anything but a .pdf, please!
Maybe that's the key to customer excitement -- how much pain they were in before you came along...
I was browsing the U3 software to see if I could find a good calendaring app, and came across Foxit for U3. Free download, nothing to lose, so I installed it and then double-clicked on a 343 page .pdf on my desktop. It opened instantly.
I gaped, I admit it. It was a jaw-dropping experience. I grabbed a coworker from the hall so she could witness it herself.
Even better was that when I did a search, I found that there is also a Foxit Reader that I could install on my regular computer, rather than the U3.
I don't have to hate .pdfs anymore. Foxit rocks!
Blake Handler has a great collection of Microsoft resources on his blog - I especially appreciate the troubleshooting section, myself. I couldn't get there a few weeks ago, and I think it might have had to do with Microsoft's changes to the MSN Spaces...
Today Microsoft converted MSN Spaces to Windows Live Spaces and hopefully this latest url (bhandler.spaces.live.com) will be permanent!