>I believe "in-place" upgrades are essential, and removing them is a >colossal mistake.
I believe so too; I would need to scrap half my test systems and I would stop testing Solaris builds on quite a few systems if in-place upgrades were killed. I believe many people will be in a similar position and I believe that Solaris quality will degrade rapidly once the ability to test on a greater variety of systems greatly diminishes. I do not see how liveupgrade is apreciably different from booting over the network and running upgrade on the alternate root. The fact that two different programs are used for those two tasks points to extremely incompetent engineering. The fact that somehow the need is felt to terminate the one where they are really both the same program running in the same environment is very short sighted. >In fact, I expect in-place upgrades to be the most common upgrade path >for many years. Space may be cheaper than it was in the past, but it's >certainly not free. Besides, scrapping in-place upgrades means >abandoning most of the installed base (including, I suspect, all machines >using pre-installed configurations). Quite. In my case, I have three systems with liveupgrade (because I need them to be on all the time) and about a dozen or so without. Liveupgrade is a burden because it requires much more maintenance (upgrade punchin twice, e.g., and remembering that you need to update certain configuration files twice) and for those systems which are not needed 24x7, it's pointless to do the additional investment. Just upgrade them overnight and be done with it. >I don't actually see why Live Upgrade is necessarily easier than an >in-place upgrade. At least with a normal in-place upgrade you're >booted off the new version in a known state, rather than an old >version of Solaris in an unknown state. I agree; in-place upgrade and liveupgrade are pretty much the same thing; liveupgrade must be harder because you can't really trust the boot environment to be the proper environment to do the upgrade in (who always updates the lu packages before doing the liveupgrade? I hardly ever do) >Going further, why isn't it possible to do an upgrade via some sort >of patch mechanism? >The argument that disks are now so large that it's no longer >necessary to worry about space simply isn't true. Systems with >18G drives (or similar) are still commonplace and entirely viable. >With the growth in virtualization, large drives will get chopped up >into much smaller chunks for allocation to installed systems. >And many Sun systems still get shipped with 73G drives. While >this may seem excessive now, it's not going to look generous >in 5 years time. Besides, Sun still sell reconditioned systems with >9G drives, and we mustn't completely ignore the hobbyist market >where less generous configuarrions are common. If Sun wants to drop the SXCR program, then abandoning in-place upgrades would be one way to diminish the effectiveness to near zero. Casper
