Community
OpenEnergyMonitor

Community

phpMyAdmin Install Issues

emonpi
Tags: #<Tag:0x00007f10adf14658>
(Dave) #1

Hi All
I am in the process of updating the firmware on my emonpi and I usually do this by provisioning a new SD card and completing an export import exercise.
I am however having issues due to various modifications that cause various errors.
One of the ways of getting around this is to edit the SQL database for the dashboards but I’m unable to install phpmyadmin as it cant find the database? below is a screenshot of the removal of the package but the issue is teh same.

Kind Regards
Dave

0 Likes

(Brian Orpin) #2

There was some discussion here about this error in a different circumstance (search on mysqld.sock found it).

One thing that has been discovered is that if you have done an upgrade-dist to an EmonSD image you will fall foul of this issue.

What is your starting point (EmonSD?), what modifications do you need and what are the resultant issues that leads you to install phpMyAdmin to fix them?

0 Likes

(Dave) #3

Hi Brian
I resolved the issue by creating a symbolic link…

ln -s /home/pi/data/mysql/mysql.sock /var/run/mysqld/mysqld.sock

I’ve never had success in upgrading and its a good mission to upgrade as so many things change as people develop, plus I’ve changed modules like the feed value to choose text size and the colours on the gauge etc.
I believe however the feed value had developed and you can choose font sizes etc.
It would be good however to be able to clone the SD card but its not something I’ve done before.
Here is my main dashboard > http://energy.m0nne.com/emoncms/dashboard/view&id=1

King Regards
Dave

0 Likes

(Brian Orpin) #4

Have you done a dist-upgrade?

Could you possibly tell me what the ProtectHome value is in file /lib/systemd/system/mariadb.service please?

If it is true, would you mind doing a small test please?

Could you change it to false (see this issue for the best way to do that) and then remove the symbolic link, reload the service and see if it solves the problem.

There might be 2 related issues here cause by this setting.

0 Likes

(Dave) #5

Protect home is false.
I’ve started again a couple of times as when I import it complains about a file not existing within the graphs code.
I haven’t done a dist-upgrade as I don’t want to break anything that requires a specific version of software.

Kind Regards
Dave

0 Likes

(Brian Orpin) #6

Ok thanks, eliminates that as a cause then.

@glyn.hudson can you think of any reason this problem is occurring (second report but slightly different circumstances). My EmonPi does not have that file.

My guess is that the /etc/mysql/my.cnf is edited during the EmonSD build process as it points to

socket=/home/pi/data/mysql/mysql.sock

Not sure why or how to fix it. This is the global file so anything using mysql would, I would have expected, used it.

0 Likes

(Dave) #7

Thanks Brian
I have just imported the backup from my working emonpi and I get the following error…

Warning : require_once(Modules/group/group_model.php): failed to open stream: No such file or directory in /var/www/emoncms/Modules/graph/widget/graph_widget.php on line 10

Fatal error : require_once(): Failed opening required ‘Modules/group/group_model.php’ (include_path=’.:/usr/share/php’) in /var/www/emoncms/Modules/graph/widget/graph_widget.php on line 10

Regards
Dave

0 Likes

(Craig Fryer) #8

I haven’t done upgrade on one of my systems. I have just tried to install phpmyAdmin and I have hit problems as well.

> E: Write error - ~LZMAFILE (28: No space left on device)
> E: Sub-process /usr/bin/dpkg returned an error code (1)

Which is a bit strange:

> Filesystem     1K-blocks    Used Available Use% Mounted on
> /dev/root        3999680 1677392   2133856  45% /
> devtmpfs          495480       0    495480   0% /dev
> tmpfs             500088       0    500088   0% /dev/shm
> tmpfs             500088   19196    480892   4% /run
> tmpfs               5120       4      5116   1% /run/lock
> tmpfs             500088       0    500088   0% /sys/fs/cgroup
> tmpfs              51200   51200         0 100% /var/log
> tmpfs              30720       4     30716   1% /tmp
> tmpfs               1024       0      1024   0% /var/tmp
> /dev/mmcblk0p3  11255794  309152  10374740   3% /home/pi/data
> /dev/mmcblk0p1     43539   22506     21034  52% /boot
> tmpfs             100016       0    100016   0% /run/user/1000

I was trying to change the Write API Key to the same as a previous install with corrupted. Otherwise I need to update about 15 devices :frowning:

Edit - formatted text for readability. BT - Moderator

0 Likes

(Bill Thomson) #9

Not too strange… Your log directory is full:

tmpfs 51200 51200 0 100% /var/log

The log directory is effectively held in RAM, (tmpfs) so a reboot will clear it.
If possible, you should try to find out why it filled to capacity, as the fact that it did get full
may be an indication of an underlying issue.

0 Likes

(Brian Orpin) #10

To where? A self install?

Neither the group and graph modules are core modules installed with Emoncms (but may be insatlled on the EmonSD - can’t check right now) and need to be installed separately from the main EmonCMS organisation. https://github.com/emoncms/graph, https://github.com/emoncms/group

No, I don’t know why either. Just how the overall system has developed over time.

0 Likes