Login, get blank page

[murdock]

Sorry, forgot to upload attachments; here they are...

[pbroste]

Hello Murdock

Here is the command that I used:

Code: Select all

ls -lahrt -R /usr/local/nag* > listnag.txt

Here is the result on the for d in `cat basedirs.txt` ; do echo "DIRECTORY $d" ; find $d -ls ; done > nag_list.txt

Thanks,
Perry

[murdock]

Hi Perry,

Please accept my apologies: I was out of the office yesterday (Mon Sep 27) and have been in meetings all day today (Tue Sep 28).

I'll have a proper response to your last message (Fri Sep 24) by tomorrow (Wed Sep 29), I just wanted to give you can update.

Rob

[pbroste]

Sounds, good please respond when you have further updates.

Thanks,
Perry

[murdock]

Hi Perry,

Detailed comparison of "your" XI (nag_list.txt) v. "my" XI (uofi_list.txt); also see attached files...

This uses a quick 'n dirty script I whipped up in a single sitting for this occasion... so "beta" / experimental. The forum software does not allow uploading scripts apparently, but renaming checkfiles to checkfiles.txt allowed it to be uploaded (so rename it back if you'd like).

There are a lot of owner and group differences, apache v. nagios; and a lot of permission differences... but again, "mine" is a new, clean fresh XI install on top of a brand new VM with CentOS patched after installing from a vanilla ISO. XI worked fine until I enabled AD integration; I did *not* manually tinker and tweak modes, owners, groups. It looks like you've got a lot of stuff installed in yours (debugging, testing I'm sure) that aren't present in mine.

report.txt has "error" codes (E***) for easy grep'ing, explained below; if there are specific directories we don't want to include in the comparison we can awk 'n sed 'em out.

Please let me know your thoughts and how you'd like to proceed...
Rob


wc -l *.txt
...
9145 nag_list.txt # "yours"
...
8999 uofi_list.txt # "mine"
...

./checkfiles nag_list.txt uofi_list.txt > report.txt
...
linecount1: 9145 (1st list)
linecount2: 8999 (2nd list)
notfound1: 567 (not in 1st list) # ~6%
notfound2: 714 (not in 2nd list) # ~8%
totalwarn: 2928 (mismatched mode, owner, group attributes) # ~32%

cut -d\ -f1 report.txt | sort | uniq -c | sort -nr | head
2179 E000 # mode differs
730 E002 # group differs
714 E1N2 # in 1st list but not in 2nd list
567 E2N1 # in 2nd list but not in 1st list
19 E001 # owner differs
...

[pbroste]

Hello @murdock

Thanks for following up with the results; thanks for diving into this and creating the script too. It appears that we ruled out permissions and verified that.

So focused on the 500 error message and the list of "internal server errors" that I totally neglected to widen the troubleshooting scope a bit more further. So with that, we realize that it is failing on the 'login.php' and want to find out what version of php you are running on that server. Very possible that could be causing the issue.

Code: Select all

php -v

Comparing I am currently running version 7.2.24, and if you want to get more php info you can run:

Code: Select all

php -r 'phpinfo();'

Thanks,
Perry

[murdock]

Hi Perry,

Attached are the results from those 2 commands...

Rob

[pbroste]

Hello @murdock

Thanks for following up, appears that the PHP is not on the latest version and want to have you run through the upgrade script by:

Code: Select all

wget -O upgrade.sh https://assets.nagios.com/downloads/nagiosxi/upgrade.sh
sh ./upgrade.sh

Please watch the results of the upgrade and let me know the results. They are also logged in '/tmp/nagiosxi/upgrade.log'

Thanks,
Perry

[murdock]

php -v # PHP 5.4.16, p2n (2v4: dev or pre-prod) -- *not* working
php -v # PHP 5.4.16, p2 (2v2: prod) -- works fine
php -v # PHP 5.4.16, p1 (1v2: test or stage) -- works fine

All 3 instances are already running PHP 5.4.16.

wget -O upgrade.sh https://assets.nagios.com/downloads/nagiosxi/upgrade.sh
sh ./upgrade.sh
php -v # still PHP 5.4.16, p2n (2v4: dev or pre-prod)

After upgrading, all 3 instances are still running PHP 5.4.16.

I hate to sound like a broken record, but the instance we're working on ("p2n") was a freshly provisioned VM with a current CentOS ISO then patched with all available updates before installing the latest XI. All 3 instances are running PHP 5.4.16; even after the upgrade p2n is still on PHP 5.4.16. p2n was working fine until I enabled AD integration -- and I've never been able to login [to the XI web GUI regardless of AD or Local account] since.

When I enabled AD integration on p2n I meticulously copied the settings to match those same settings on p2 and p1.

I bet if there were a way to properly remove AD integration from the CLI that the web GUI on p2n would start working again.

As always, thank you for looking in to this,
Rob

[murdock]

PS, sorry, I forgot to attach upgrade.log; here it is now.