Fresh install fails to boot --with a "transaction is destructive" error in console???


Decided to switch my install of TW over to Leap 15.1 . . . checksum checks out . . . ran the install (twice) using “expert partitioner” and ran the install, hoping to pick up my previous /home directory from the TW install?? At the beginning of the installs an error showed up saying, “efi partition is not larger than 256 MB and might not boot, do you want to continue?” And I said yes . . . and we continued; in the last 5 minutes of the install another error window opened showing that “online repositories could not be accessed” and “5 packages failed to load” . . . some parts of “yast” . . . options to choose “retry” “abort” or “skip” . . . after several “retries” failed I went to “skip” . . . and the installation “finished.”

On reboot it went to the GRUB menu showing up as “GNU LINUX"which went to the “Leap” splash” page and then went to a TTY to “emergency mode” . . . giving me a “log in” cursor, log in didn’t accept my user name and password. I shut it down and rebooted into the install DVD and selected “rescue system” . . . went through the dmesg stuff?? again wouldn’t log in or figure it out. I restarted and tried the second install . . . same result . . . ???

Is the potential reason for “failure” due to the EFI partition (set by OSX at 200 MB) not being large enough? Or, is it because the Leap & TW permissions are not compatible, so I can’t re-use the user name directory from the TW install over in Leap?? Or, is this a “Beta” problem? Installing in a '12 Mac Pro . . . multi-boot system . . . number of linux installs and OSX installs in 3 internal HDD/SSD . . . .

Thanks for any clarifications,


No, that should not be a problem, unless the EFI partition is full.

Or, is it because the Leap & TW permissions are not compatible, so I can’t re-use the user name directory from the TW install over in Leap??


There may be incompatibilities. But those would not show up until a user attempts to login.

Yes, it might be a problem using a 12 Mac Pro. But I have no experience with those, so I don’t know whether that would be a problem.

When did this “transaction is destructive” message show up? Did you get a grub2 boot message? Did it seem to be loading the kernel?

Which MacOS version do you have? This smells like it could be SIP-related. Multiboot on my iMac gave me fits. ISTR disabling SIP is a must.


Perhaps I should have said, “won’t log in” . . . “transaction is destructive” shows up after the Grub list shows up, selecting the Grub-GNU item, the Leap Splash shows up, then it went to TTY window, which wouldn’t accept my user log in and password that I had just set up . . . it showed the “emergency log in” in the TTY, then the “transaction is destruction” . . . perhaps after trying to “log in” to the TTY session?

I don’t know if the EFI is “full” or not . . . the GRUB file is usually a few KB . . .


I’m pretty sure that SIP is disabled . . . I have installed other linux installs that went well since my last OSX update . . . they went well . . . . I have found the OpenSUSE installer to be a little more “finicky” than the ubuntu installer . . . .

“transaction is destructive” suggests maybe you should give us a look at the output from parted -l.

At the end of an installation it is possible to pause instead of proceeding to reboot. Once paused one can Ctrl-Alt-F2 to capture logs by saving them to a disk partition or USB stick. You can save a capture of an error message screen in similar manner. It’s not a terribly difficult process, and probably would be the only possible way for anyone to discover what happened. Other ttys listed when you first switch to F2 might already have clues showing. The logs are the same ones available if the installation sufficiently completes that it can be booted into, even if doing so via OM or USB stick is required.

The EFI partition size warning is an unfortunate waste of people’s time. AIUI, the only time it might matter is if someone is using a storage device of at least 16TB (or much more?) with 4k sectors both logically and physically, in which case the required minimum size would be ~256MB.

Boot with plymouth.enable=0 on kernel command line, it will provide better feedback what happens.

then it went to TTY window, which wouldn’t accept my user log in and password that I had just set up . . . it showed the “emergency log in” in the TTY

At this point it waits for root password, it does not ask you for user name.

then the “transaction is destruction” . . . perhaps after trying to “log in” to the TTY session?

So your summary is misleading. Boot does not fail with “transaction destructive”, this message is result of some activity after boot has already failed and so is irrelevant for troubleshooting.

The only reason systemd goes into emergency login mode is failure to mount mandatory filesystem (i.e. one defined in /etc/fstab and not having “nofail” option). Check what you have there and what is currently mounted when you login into emergency console.

@mrmazda: Thanks for the follow up, appreciate that . . . problem is the system won’t let me log in via TTY . . . even using “rescue” I can’t log in . . . .

@arvidjaar: Yes, “summary is misleading” . . . after I posted the thread I tried to edit it to say “fails to boot/log in . . .” but couldn’t figure out how to get that done . . . it is what it is . . . the system “boots” but doesn’t show log in GUI and/or doesn’t let me log into TTY shell . . . .

So, same answer to you on the suggestion to “boot with plymouth.enable=0” . . . after the first install I could get to # user, but console was “unresponsive” . . . using “zypper ref” . . . response was “command not found” . . . and so forth, “reboot” brought a new cursor line only, etc. The console appears to be “unresponsive” to command input . . . .

I did take some “screen shots” using my iPod to capture some of the errors . . . one of the shots shows error:

Failed to start Transaction is destructive.

And, in one of the “emergency” or “rescue” modes it suggested checking something, possibly “systemctl”?? which showed something like what in ubuntu is called “dmesg”??? long list of many, many lines of data, which I scrolled through to one place that showed the text in Bold font, which I assumed to be an “error” . . .

DMAR-IR: This system BIOS has enabled interrupt remapping on a chipset that contains an erratum making that feature unstable.  To maintain system stability interrupt remapping is being disbaled. Please contact your BIOS vendor for an update.


Don’t know if that is “relevant” or not, as the Mac install is using EFI . . . .


Pretty sure might not be good enough. AIUI, SIP isn’t necessarily persistent, but can be silently restored by certain MacOS updates, such as the BIOS.


Sure, anything is possible . . . haven’t done any OSX updates lately; but earlier yesterday I did a Debian based install on the same computer in the same HDD . . . all went well with that install . . . . A little while later, running the Leap install, so, mentally warmed up for installations after the earlier “reminder” . . . all seemingly did not go well . . . ??? Relatively “new” to OpenSUSE, but not new to linux installations . . . some things often are “pilot error” . . . but, sometimes it goes the other way . . . .



No need for further discussion . . . issue didn’t seem to be “fix-able” via hints offered . . . moved on to the “nuke” phase of making the adjustment to the problem . . . . I’ll try again in a few days to find another iso . . . see if that makes any difference.

I’m attributing the “issues” to the “incomplete” download/install of the system . . . coming up “5 packages” short that the installer “could not retrieve from repositories”??? Or somehow something in Grub was not lining up??

Leap 15.1 system is being “dissolved” back into the ethers as we speak . . . failed install(s).


Just to be clear to people reading this – Leap 15.1 is doing fine here, both with real machines and virtual machines.

To the OP for this thread: I’m sorry to hear that you are still having problems.


Yes, still having problems . . . something about that install(s) did not go “well” . . . and then after I posted the last post this morning, even trying to “erase” the Leap partition in OSX to get it nice and clean for the next install, using Disk Utility (as I did before the Leap install) to lay a bunch of “zeroes” . . . has not been able to be done . . . “not enough space to complete the operation” . . . so “not enough space to erase the partition” . . . ???

And, then, something seems to have been distorted in Grub, as trying to do an open reboot into the Grub list . . . brought a Grub “rescue” page . . . . I’m typing this now in Gecko to see if using that OpenSUSE partition/system will be able to get things back to normal . . . . It’s just “odd” that trying to run an install in what was a “Tumbleweed” system partition seemed to go “sideways” . . . and then leave some “distortion” in its tracks.

Couldn’t seem to get any insight from the guru’s here about what the “issue(s)” could be, to try to fix it from what was installed . . . sometimes installs go “off track” . . . this seems to have been one of them . . . reasons unknown . . . . Appears that “transaction is destructive” . . . was . . . . It won’t be today, but if and when I get another .iso ready and run the install . . . I’ll post back on it.

Unfortunately, I am unable to look over your shoulder to see what is going wrong. I’m not seeing enough information to know what is broken. And I’m not seeing enough to know what questions I should be asking to get more information.

For now, I’ll continue following the thread and hope that something comes up that I can recognize as pointing to the problem.


OK, made the time, fresh download, freshly burned to usb drive . . . ran the install and it’s the “third time” situation . . . it was “pilot error” as I figured, but I just didn’t know the “where” of it . . . but I also think the installer could have done a “better” job . . . as well.

The clue was in that “dmesg error” that I posted where it was referring to a problem with the “BIOS” . . . something, something . . . . In today’s install when I went through the “expert partitioner” mode in the bootloader section the “auto-filled” choice at mid-lower left window was “BIOS Boot” . . . and in the other installs I didn’t catch that one, since the EFI partition is usually being used by Grub . . . But . . . also because I booted the “EFI” version of the Leap iso . . . so . . . seemingly the installer should “know” that I’m wanting to do an “EFI” install . . . conversely, it seems to catch the other choices, it loaded the “/” choice for system partition, and it auto-fills the “/home” and recognizes one of my “swap” partitions . . . but, seemingly “missed” the “EFI” part and auto-set it as “BIOS” . . . today, I saw that, and changed it to “EFI system” . . . . And, all went well . . . hooked up the TW user directory . . . and more or less it seems to be working . . . “Gmail” seems to think I’m “hacking” my email account, but otherwise . . . we are logged into Leap 15.1 and it is working to make this post . . . .

Only thing about this forum that is different than ubuntu, is there it is possible to edit the subject line via the first post . . . or “mark as solved” but here it isn’t obvious how to do that.


That’s certainly an improvement. I’m glad to see that.

“Gmail” seems to think I’m “hacking” my email account, but otherwise . . . we are logged into Leap 15.1 and it is working to make this post . . . .

You can maybe start another thread on that. I’m pretty sure that there are people who have gone through that problem.

Yes, me too. OK, on the “start another thread” on the Gmail issue . . . it said something about “network protocol failure” . . . again with the “failure” . . . I’m over in an old Chromebook and gmail was OK . . . .

It does look like there is the “Gmail” thing, and also seems like Leap didn’t catch the TW user “font sharpness” settings . . . fonts were almost unreadable . . . . I shut the Mac down, and those issues will be for another day and I guess another sub-forum.