Strange orbiter error in web interface

[wortmanb]

I've just started getting this error when navigating to the orbiter page in pluto-admin:

Fatal error: mysql error: [1030: Got error 28 from storage engine] in EXECUTE(" SELECT IF(FK_DeviceCategory=2,'mobile_orbiters',IF(Device.FK_DeviceTemplate=62,'on_screen_orbiters','standard_roaming_orbiters')) AS OrbiterGroup, Device.*, DeviceTemplate.Description AS TemplateName, DeviceCategory.Description AS CategoryName, FK_DeviceCategory, Manufacturer.Description AS ManufacturerName, IsIPBased, RegenInProgress,RegenStatus,RegenPercent, DeviceData.Description AS ddDescription, ParameterType.Description AS typeParam, Device_DeviceData.IK_DeviceData, Device_DeviceData.FK_DeviceData, ShowInWizard, ShortDescription, AllowedToModify, DeviceTemplate_DeviceData.Description AS Tooltip FROM Device INNER JOIN DeviceTemplate ON Device.FK_DeviceTemplate=PK_DeviceTemplate INNER JOIN DeviceCategory ON FK_DeviceCategory=PK_DeviceCategory INNER JOIN Manufacturer ON FK_Manufacturer=PK_Manufacturer L in /var/www/pluto-admin/include/adodb/adodb-errorhandler.inc.php on line 77

I noticed that my remote orbiter on my laptop suddenly stopped working and my core had trouble rebooting.  It was when I went to check it on the admin site that I discovered the above problem.

I've no idea what happened to corrupt my database, but is there any easy way to fix this?  Would upgrading to 0710 take care of it (I'm running 0704 right now)?

Thanks!


Bret Wortman

[Zaerc]

So the installation was already corrupt before you started the update?  I think you forgot to mention that in your other posts about the update failing...

[wortmanb]

I didn't actually mean to start the update.  I ran update-prepare.sh (or whatever the script name was; it ended in prepare) and that was all.  When I rebooted the core, it just started the process on its own.  Maybe that's as intended, but given the lack of documentation for the upgrade package, the "prepare" file led me to believe (and a cursory read of the script itself confirmed) that it would just download the needed files.  I had thought it'd be up to me to start the actual upgrade at a convenient time.

I may have done this to myself, but it wasn't at all my intention.  And I'm surprised that the scripts would have started modifying files without first ensuring that all the downloads were in place and intact!

(That's not a slam, just a slight critique).


Bret

[Zaerc]

And you don't think this "minor detail" was worth mentioning in the other posts you made when asking for help with the upgrade that supposedly failed for no apparent reason???

Maybe you can find some other random system scripts to run manually, and who knows they might even restore everything again.  

[wortmanb]

Just to close the topic --

1.  I logged on via a Kubuntu live CD.
2.  I gathered a list of all the .pbackup files on the system and moved every one of them to the non-pbackup name:
     # find . -name "*.pbackup" > ~/restore.sh
     modified the above file so that each line was of the form:  mv file.pbackup file
     executed the above file.
3.  This allowed the box to reboot, and only minimal intervention was required to restore the network configuration back to where it worked again.

[wortmanb]

By the way, there was no problem in my database -- but one of the scripts that produced the page for the admin interface had been bollixed by the crashed install and thus I was seeing what was referenced below.  Everything now works as intended and I'm going to seriously rethink my approach to upgrading this time.

So the "minor detail" wasn't in any way the cause of the problem but was, in fact, another symptom.  So you're right, I should have mentioned it.  But it wasn't the cause....

As for running random scripts, I ran the prepare script, which in the case of an Internet download has no code that refers to rebooting to continue the upgrade -- that code is under the DVD branch of the if.  I naturally assumed I'd have to take manual steps to start the upgrade.


Bret

[Zaerc]

By the way, there was no problem in my database -- but one of the scripts that produced the page for the admin interface had been bollixed by the crashed install and thus I was seeing what was referenced below.  Everything now works as intended and I'm going to seriously rethink my approach to upgrading this time.

So the "minor detail" wasn't in any way the cause of the problem but was, in fact, another symptom.  So you're right, I should have mentioned it.  But it wasn't the cause....

As for running random scripts, I ran the prepare script, which in the case of an Internet download has no code that refers to rebooting to continue the upgrade -- that code is under the DVD branch of the if.  I naturally assumed I'd have to take manual steps to start the upgrade.


Bret

Right, so the failure of the update had nothing to do with your system being corrupt before starting the update?

[wortmanb]

The update caused my system to be corrupt.  Did you read that last post or are you just so convinced that I must be an ignoramus that you can't be bothered?

Manually reverting everything back to the pre-upgrade state has resulted in a perfectly functioning 0704 system again.  I repeat:  THE UPGRADE CAUSED MY PROBLEM, IT DIDN'T RESULT FROM A CORRUPT SYSTEM!

[Zaerc]

The update caused my system to be corrupt.  Did you read that last post or are you just so convinced that I must be an ignoramus that you can't be bothered?

Manually reverting everything back to the pre-upgrade state has resulted in a perfectly functioning 0704 system again.  I repeat:  THE UPGRADE CAUSED MY PROBLEM, IT DIDN'T RESULT FROM A CORRUPT SYSTEM!

I remember specificly asking you, but somehow you couldn't be bothered to give a straight answer to a simple question.  And going over a few of your previous posts, it seems pretty clear that you most likely caused these problems yourself to begin with by running system scripts manually and manipulating the database directly. 

Just FYI, I didn't think you were an "ignoramus" before, but I do now.  And just in case you hadn't noticed, your caps-lock seems to be stuck.

[wortmanb]

You know, I'm taking this much too personally.  Sorry about that -- I guess having a crashed core got me a bit emotional.  Mea culpa.

I didn't touch the database manually.  I ran one script, upgrade-prepare.sh.  That's it.  Everything that followed happened as a result of running that one script.

As for the shout, I guess I wasn't getting my own point across very well -- that this one action caused everything which followed and I was simply asking for guidance in how to best proceed from those who have more linuxmce experience than me rather than plunging in blindly and causing more problems.

The only mucking about I did manually were the steps I took to restore things back the way they were.

So, how is one supposed to perform an upgrade using the gutsy-upgrade-scripts package?

[Zaerc]

I suppose that makes two of us.  I'll take back that "ignoramus" remark, that was a stupid and hurtful thing to say.

...
So, how is one supposed to perform an upgrade using the gutsy-upgrade-scripts package?

I hate to admit it, but I have no idea as I always install fresh.