Deploy task remains...

[Wayne Workman]

r7142

Attempting to deploy an image that I’m playing with the partition resizing on, getting a lot of failures. I’m just playing around with the d1.partitions file. But anyways, when I deploy the task to the computer, the computer tries to image and fails, and then clears the task in the DB so that it no-longer shows inside of Task Management at all - however the task remains.

The activity pie chart shows 1 task active, and when looking at task management, the task very, very briefly shows and then disappears. It kinda looks like a JavaScript issue.

After failure, here is what is in the tasks table:

MariaDB [fog]> select * from tasks;
+--------+-------------+---------------------+---------------------+------------+-------------+-------------+-------------+--------------+-----------+------------------------+------------+------------+---------+-----------------+-------------------+----------------+-----------------+---------------+----------------+-----------------+-----------------+------------------+---------+---------------+--------------+
| taskID | taskName    | taskCreateTime      | taskCheckIn         | taskHostID | taskImageID | taskStateID | taskIsDebug | taskCreateBy | taskForce | taskScheduledStartTime | taskTypeID | taskPCT    | taskBPM | taskTimeElapsed | taskTimeRemaining | taskDataCopied | taskPercentText | taskDataTotal | taskNFSGroupID | taskNFSMemberID | taskNFSFailures | taskLastMemberID | taskWOL | taskPassreset | taskShutdown |
+--------+-------------+---------------------+---------------------+------------+-------------+-------------+-------------+--------------+-----------+------------------------+------------+------------+---------+-----------------+-------------------+----------------+-----------------+---------------+----------------+-----------------+-----------------+------------------+---------+---------------+--------------+
|      1 | Deploy Task | 2016-04-11 22:04:22 | 2016-04-11 22:04:23 |          1 |           3 |           3 |           0 | wayne        | 0         | 0000-00-00 00:00:00    |          1 | 0000000000 |         |                 |                   |                |                 |               |              1 |               1 |                 |                0 | 1       |               |              |
+--------+-------------+---------------------+---------------------+------------+-------------+-------------+-------------+--------------+-----------+------------------------+------------+------------+---------+-----------------+-------------------+----------------+-----------------+---------------+----------------+-----------------+-----------------+------------------+---------+---------------+--------------+
1 row in set (0.00 sec)

However, when I browse to boot.php using the computer’s MAC to see what’s being given as an iPXE boot script, this is what I see.

#!ipxe
set fog-ip 10.0.0.4
set fog-webroot fog
set boot-url http://${fog-ip}/${fog-webroot}
kernel bzImage32 loglevel=4 initrd=init_32.xz root=/dev/ram0 rw ramdisk_size=127000 keymap= web=10.0.0.4/fog/ consoleblank=0 rootfstype=ext4 rootfstype=ext4 mac=14:fe:b5:df:fc:7e ftp=10.0.0.4 storage=10.0.0.4:/images/ storageip=10.0.0.4 web=10.0.0.4/fog/ osid=50 consoleblank=0 irqpoll hostname=LIVINGROOM chkdsk=0 img=Fedora23LivingRoom imgType=n imgPartitionType=all imgid=3 imgFormat= PIGZ_COMP=-6 hostearly=1 type=down
imgfetch init_32.xz
boot

Right now, I’m clearing the task just by truncating the tasks table.

[Tom Elliott]

Do, by chance, have a scheduled task for this host?

[Tom Elliott]

I ask this because I just tested, both an upload and download tasking and neither one “remained”.

That said, I do have a scheduled wake-up task (which for me doesn’t do anything, but it was just a tester to make sure things are still operating in regards to task scheduler and what not.).

The worry part, to me, is the fact that it shows as “in progress” with the state ID (3). I’m just guessing that either the host was in the middle of tasking, or something else went horribly wrong.

[Wayne Workman]

@Tom-Elliott I don’t have any scheduled tasks at home. And something did go horribly wrong lol. Partclone threw some really nasty errors while I was fiddling with the d1.partitions file trying to get my image to expand the way I wanted.

As far as I can tell, this only happens when a deploy task fails. The task is just stuck in limbo on the task management page. It’s there briefly, then totally disappears.

[Tom Elliott]

@Wayne-Workman It is fixed in the latest, allowing you to cancel the task properly.

[Wayne Workman]

I believe this was the same bug here, same version too, 7142, so I’m linking the two together.
https://forums.fogproject.org/topic/7144/build-7142-task-management-broken