Lubuntu 20.10 calamares "failed to create a partition table"

[Chris Guiver]

This really is a re-filing of https://bugs.launchpad.net/ubuntu/+source/calamares/+bug/1851188

That bug report was reported as 'Fix Released' (for a number of days the issue WAS unable to be re-created), but every so often daily builds had it re-occur (then occur for days..).

I did this identical QA-test install using yesterday's daily build and had NO issues, today it crashes.

dell [optiplex] 755 (c2d-e8300, 8gb, amd/ati radeon rv610/radeon hd2400 pro/xt)

Session log will be attached, I'll re-run the install with calamares verbose logging (which sometimes fails, sometimes passes.. but hasn't been run/attempted yet).

ProblemType: Bug
DistroRelease: Ubuntu 20.10
Package: calamares 3.2.20-0ubuntu1
ProcVersionSignature: Ubuntu 5.4.0-26.30-generic 5.4.30
Uname: Linux 5.4.0-26-generic x86_64
.etc.calamares.modules.finished.conf:
 ---
 restartNowMode: user-checked
 restartNowCommand: "systemctl -i reboot"
.etc.calamares.modules.partition.conf:
 efiSystemPartition: "/boot/efi"
 enableLuksAutomatedPartitioning: true
 userSwapChoices: none
 drawNestedPartitions: true
 defaultFileSystemType: "ext4"
.etc.calamares.modules.shellprocess_logs.conf:
 ---
 dontChroot: true
 timeout: 30
 script:
     - calamares-logs-helper @@ROOT@@
.etc.calamares.modules.unpackfs.conf:
 ---
 unpack:
     - source: "/cdrom/casper/filesystem.squashfs"
         sourcefs: "squashfs"
         destination: ""
ApportVersion: 2.20.11-0ubuntu33
Architecture: amd64
CasperMD5CheckMismatches: ./boot/grub/efi.img
CasperMD5CheckResult: fail
CasperVersion: 1.446
CurrentDesktop: LXQt
Date: Sat May 9 21:20:20 2020
LiveMediaBuild: Lubuntu 20.10 "Groovy Gorilla" - Alpha amd64 (20200508)
RelatedPackageVersions:
 calamares-settings-ubuntu-common 1:20.04.2
 calamares-settings-lubuntu 1:20.04.2
 xfsprogs 5.6.0-1ubuntu1
 btrfs-progs 5.6-1
SourcePackage: calamares
UpgradeStatus: No upgrade log present (probably fresh install)

--- Reasoning for this (taken from QA-test comment)
dell [optiplex] 755 (c2d-e8300, 8gb, amd/ati radeon rv610/radeon hd2400 pro/xt)
testcase: full disk, bios, no encryption, internet
I don't see need to do installs this early, however this is for purposes of https://bugs.launchpad.net/ubuntu/+source/calamares/+bug/1851188 brought up by https://discourse.lubuntu.me/t/error-install-20-04/1087 & it still occurring (status is 'fix released' yet this box had the issue repeat with final RC for 20.04) I didn't get response on IRC over changing "fix released" (weeks ago, twice) so this will be a new bug if it re-occurs, or will need to be marked duplicate of old bug... race bugs like this with many stack issues are a pain :(

[Chris Guiver (guiverc)]

This is a ~exact re-run of the QA-test that was used for led to the creation of this bug report.
I booted same box, changed screens to match my setup (right display is portrait-left tilted) as done last test

Then opened terminal and ran `sudo -E calamares -d` INSTEAD of clicking install icon on desktop

FYI: Calamares will show being loaded twice; this was my error, as opened at terminal first time, it displayed on my portrait-left tilted display and was unusable (buttons were off screen), I was unable to move window to other display so exited.... This behavior has been noted before, maybe requires bug report, but I'm ignoring that currently... Mentioned here only to explain why calamares was run twice in log (opened both times with same command)

With the run of `calamares -d` instead, the install was flawless (I haven't rebooted & tested yet, but I'm betting it is).

This is the log - ~/.cache/calamares/session.log

[Ubuntu QA Website (ubuntuqa)]

This bug has been reported on the Ubuntu ISO testing tracker.

A list of all reports related to this bug can be found here:
http://iso.qa.ubuntu.com/qatracker/reports/bugs/1877733

[Chris Guiver (guiverc)]

Another try again on same box (~day 3 I believe)
http://iso.qa.ubuntu.com/qatracker/milestones/413/builds/212115/testcases/1701/results

No issue today... But Leok tried on box the original report was filed on (https://bugs.launchpad.net/ubuntu/+source/calamares/+bug/1851188) and didn't have an issue either... Do we get fails/passes on same ISO/days?

[Chris Guiver (guiverc)]

Another try again on same box (~day 4 I believe)
fails today :(

lubuntu@lubuntu:~$ apt-cache policy calamares
calamares:
  Installed: 3.2.20-0ubuntu1
  Candidate: 3.2.20-0ubuntu1
  Version table:
 *** 3.2.20-0ubuntu1 500
        500 http://archive.ubuntu.com/ubuntu groovy/universe amd64 Packages
        100 /var/lib/dpkg/status

[Chris Guiver (guiverc)]

I tried it again without rebooting (usually I reboot)
Calamares gave the same results on screen

[Chris Guiver (guiverc)]

Still no reboot, re-try during same session as #5 & #6
`sudo -E calamares -d`
It worked this time :)
(session.log will have tests for #5 & #6 comments I believe)

[Chris Guiver (guiverc)]

Testing on same box; day 5 ISO
FAILED again today :(

[Chris Guiver (guiverc)]

Repeat (day 5) of comment #8 via `sudo -E calamares -D` staring calamares
this install worked okay (to reboot stage)

[Leó Kolbeinsson (leok)]

Just tested this again on the same machine that the original bug 1851188 occurred on:

Acer [Aspire] E3-111-P60S (Pent.N3530, 4GB, Intel HD Graphics, Realtek RTL8111/81681/8411 GB Ethernet, Qualcomm Atheros AR9462 Wireless, Bluetooth Atheros A315-53, 500 GB hd)

The install was successful with no errors. ( this was installed using the calamares GUI - i.e normal install).

[Chris Guiver (guiverc)]

@Leok suggested
- i boot ISO
- wait 15 secs before clicking install (I waited 15-20 secs according to wall clock)

otherwise IDENTICAL options chosen (same username/hostname/passwd etc) as comment #8
and install worked this time !!??

[Chris Guiver (guiverc)]

Re-test of comment #11 (same d755-8 box)

ISO booted, waited 20-21 secs after boot for machine to 'settle' then started. install failed.

Walked to main machine (another room) & recorded failure on iso.qa.ubu..
Returned & closed 'calamares' (still on screen so error message remained)
waited a bit then restarted install

Second time install was flawless??

[Leó Kolbeinsson (leok)]

Tested once again on the machine that the original bug 1851188 occured on:

Acer [Aspire] E3-111-P60S (Pent.N3530, 4GB, Intel HD Graphics, Realtek RTL8111/81681/8411 GB Ethernet, Qualcomm Atheros AR9462 Wireless, Bluetooth Atheros A315-53, 500 GB hd)
Testcase:BIOS,no encryption.full disk,internet

Normal install looking for bug 1851188 or 1877733 this running on the machine that encountered the original bug report 1851188

The installation failed (unable to create partition table) the installer was then rerun without rebooting and installed the system with no errors .
Rebooted-logged in and all normal.

As the original attempt was unsuccessful the result "failed" was selected. The session log has been attached to this comment.

[Chris Guiver (guiverc)]

I have just completed three QA-installs (groovy daily, 2020-07-24) on
hp dc7700 (c2d-e6320, 5gb, nvidia quadro nvs 290)
recorded at http://iso.qa.ubuntu.com/qatracker/milestones/414/builds/217793/testcases/1701/results

Only the first install failed with this issue.

First install had whatever prior installs I had on box, which included (I think) 3 partitions, original (full disk) install, followed by two install.alongside (or install.alongside + manual.partitioning).

To get around issue, I used `sudo -E calamares -d` and install worked.

The subsequent installs were full disk over a simpler layout (single partition though 2x encrypted partitions) and worked flawlessly.

[Chris Guiver (guiverc)]

Lubuntu groovy daily [2020-10-19] on
- dell [optiplex] 755 (c2d-e8300, 8gb, amd/ati radeon rv610/radeon hd2400 pro/xt)

testcase: full disk, no encryption, BIOS, internet

FAILED with this issue.

I exited installer, restarted `calamares` & tried the install again, it worked flawlessly second time.

alas I forgot to save the session.log sorry

[ԜаӀtеr Ⅼарсһуnѕkі (wxl)]

Am I understanding correctly that with one particular testcase the initial run fails, but further runs succeed?

Can this be reproduced on other machines and/or virtual machines?

[Chris Guiver (guiverc)]

No I've not had this issue occur on another box.
It only very rarely occurs on this box anyway

(I don't know how to query http://iso.qa.ubuntu.com/qatracker/api to download all my tests so I can search there to get some accurate stats, no reply on ubu forums requests for help)

It's my belief this is still the race condition of https://bugs.launchpad.net/ubuntu/+source/calamares/+bug/1851188 that never left.. just rarely shows itself..

It's happening only very rarely.

[ԜаӀtеr Ⅼарсһуnѕkі (wxl)]

Harumph. Well, I guess we'll call this incomplete until we see some patterns emerge, but let's keep it in the back of our minds.

[Leó Kolbeinsson (leok)]

@wxl @guiverc

Just for the record:
I have not run across this issue again since 13.05.2020 see comment #13 - that occurred on ISO from the same day.

[Launchpad Janitor (janitor)]

[Expired for calamares (Ubuntu) because there has been no activity for 60 days.]