GUI update reporting failed on nanoBSD but update keeps running and works
-
On my Alix 2D13 running 2.3.1-DEVELOPMENT I go to System->Update, it finds an update and I press the button to confirm. I get initial output shown about updating the repository catalogue, then, at about the point where the "Update from console" would then ask to confirm to proceed with the upgrade, the GUI reports that the upgrade has failed.
Actually the upgrade seems to be proceeding OK in the background. If I go to the console and use option 13, it tells me that an upgrade is already running. After a while the system reboots and is running the new version.
So it seems there is just some problem with the GUI understanding the upgrade process - it is prematurely thinking that the upgrade process has failed.
When the next snapshot has built, I will try using the GUI to update my test VMs, to see if the problem is just with nanoBSD, or more general.
-
Not seeing that on a number of full installs I've upgraded in the past few hours. I don't have any nanobsd systems handy at the moment (I'm in Germany this week), but will bring up some test VMs and see.
-
-
Thanks, checking into it.
-
i think its got to do with the update display page, earlier when u clicked update in menu, it used do all the checking and comparing then once done display page but now in the new style, the page is displayed with a rotating wheel which then goes away once it populates the versions and then when u click update the page needs to keep refreshing and below data keep appending but soon it stops and says failed, could be its checking some variable etc to find out if update is still running or no
-
I had one a few hours ago where I clicked to confirm, then a heap of text output scrolled through the window, then some of the genuine expected first output about updating the repository catalogue. The page went to saying the update was successful (in about less than 1 second!) and started the countdown from 90 for reloading the page (where it expects that the system is rebooting). After the countdown it then refreshed and went back to the dashboard, which was still there of course on a system that was still up.
Some minutes later the system rebooted and came up with the new version. So the real update was still going on in the background and worked fine.
This was the same system that I reported on at the start of this thread, but after the next build became available.
The heap of text output looked (in shape…) like upgrade type of messages. So maybe there is also some way for the GUI front-end to see an old upgrade log, display and parse it, and thus think that the upgrade is done and ready for a reboot.
It was all so quick and I wasn't expecting that, so I have no recording of it.
-
Looks like pkg's json output is failing or misbehaving there for some reason, though pkg is continuing to function just fine. Not seeing that here on a few tests though, on both 32 and 64 bit.
Everything I have readily available until I'm back home Friday is much faster than what you're probably running with nano though. Might be something only replicable with slower CPUs.
-
actually few days back i also had that successful message and then 90secs later back on dashboard but it was updating in the background but now since the past 2 days or so i keep getting failed but it still updates in the background
-
Updated an ALIX with NanoBSD today and it went smoothly, no problems I could see.
-
-