| Author |
|
MrGibbage
Super User
Joined: October 23 2006
Location: United States
Online Status: Offline
Posts: 513
|
| Posted: April 04 2015 at 18:14 | IP Logged
|
|
|
I waited until the time felt right to do the upgrade. I
have been hanging out on 2.1.4 and today I installed a
new KPL 2334-222 and I knew that to install it, I would
need to upgrade PH. So away I went. I renamed the old PH
directory, uninstalled the old versions using the windows
control panel uninstall program feature. I installed the
new 2.1.5c version, copied over the old pwrhome.ini and
db files and rand the phupg utility (no errors).
Powerhome 2.1.5c starts up but as soon as I start
interacting with it, it freezes up and it crashes. I have
tried rebooting a few times, but still I get crashes.
Oh well, I figured. I'll just go back to 2.1.4. Nope. Now
I get an error about a missing "PBVM115.dll".
Help me, Obi-Wan Davenobi! You're my only help! Where do
I start with troubleshooting this?
System info
Windows Vista
Service Pack 2
x64
This is a standalone PH box and doesn't do anything else
other than run PH.
__________________
Skip
|
| Back to Top |
|
| |
dhoward
Admin Group
Joined: June 29 2001
Location: United States
Online Status: Offline
Posts: 4447
|
| Posted: April 04 2015 at 19:07 | IP Logged
|
|
|
Skip,
Uninstalling the old versions is probably what is
keeping you from going back to 2.1.4. You should be
able to do a full install of 2.1.4 (renaming any
existing directories...you never want to install into
an existing dir) and then copy your backed up (or
original) 2.1.4 db and ini and you should be good to
go.
However, it would be best to stay on 2.1.5c if you
can. Do you get any error messages or does PowerHome
just "hang"? What directory is 2.1.5c installed in?
Give me a little info about the freeze and crash.
Also, you may want to try reinstalling 2.1.5c in case
the install was corrupted (just rename your existing
2.1.5c directory...don't uninstall anything).
Let me know,
Dave.
|
| Back to Top |
|
| |
MrGibbage
Super User
Joined: October 23 2006
Location: United States
Online Status: Offline
Posts: 513
|
| Posted: April 04 2015 at 19:30 | IP Logged
|
|
|
The only message is the standard windows "%APPNAME% is
not responding" and windows can check for a solution
(like that EVER helped!), with the options to check for a
solution, close the program or to wait for the program to
respond.
I wonder if it could be a plugin or controller problem. I
remember I had at least one bad controller (or maybe it
was a plugin) but 2.1.4 was happy to just ignore it. It's
a plugin for my Denon receiver. I could just delete it,
but I can't interact with PH at all. It crashes so
quickly.
I do have PH installed in the standard c:\powerhome
folder.
Is there a way to selective see and choose what
controller and plugins are loaded?
Oh wow, just now, while I was typing this, I was watching
the PH screen. I had that dialog up regard the app crash.
Well, it just went away, so it seems to be responding.
For now.
I went to help-about and I see four controllers loaded
Elk
USB_UIRT1
USB_UIRT2 (I have two of them)
PLM
That all looks good
Under plugins, I do have one invalid plugin
PowerHome Plugin ID: DENON
Plugin failed during initialization.
So, let's see if I can open the Insteon Explorer...
Check! It worked.
OK, I think I need to look into that plugin. How do I get
rid of that and test some more?
Warning: Standby for questions regarding the KPL if I get
this working :)
__________________
Skip
|
| Back to Top |
|
| |
MrGibbage
Super User
Joined: October 23 2006
Location: United States
Online Status: Offline
Posts: 513
|
| Posted: April 04 2015 at 19:36 | IP Logged
|
|
|
Well, that didn't last long. It froze up again. I'll let it
sit on the frozen screen (it is an all white screen and the
title bar says "not responding")
And of course, while I was just typing this, it went away
and started responding again. So it was frozen for about a
minute or two.
__________________
Skip
|
| Back to Top |
|
| |
MrGibbage
Super User
Joined: October 23 2006
Location: United States
Online Status: Offline
Posts: 513
|
| Posted: April 04 2015 at 19:47 | IP Logged
|
|
|
Sorry for all of the rapid-fire responses. But I have
more information.
If I just let it sit and don't touch it, periodically
(live every couple of minutes) I will get a popup that
says "Lost communications with M1XEP on Elk M1 Controller
ELK. Attempting to reconnect. And it is supposed to count
down 3, 2, 1 and then close by itself, right? Well, it is
stuck on 3. And I can't click it or anything else. But if
I wait long enough, it will eventually go away, but I
still can't do anything with it. Not Responding. But if I
wait long enough, eventually it will start working again,
like nothing was wrong.
I know the ethernet controller is fine. It was working
this morning on 2.1.4, and the Elk is doing fine too. So
it's something else.
__________________
Skip
|
| Back to Top |
|
| |
MrGibbage
Super User
Joined: October 23 2006
Location: United States
Online Status: Offline
Posts: 513
|
| Posted: April 04 2015 at 19:49 | IP Logged
|
|
|
By the way, every time I post something in the forum here,
I get an error
error '8004020f'
/forum/functions/functions_send_mail.asp, line 151
And when I press the back button, sure enough the post did
indeed post. But something is wrong. This has happened
every time I posted today.
__________________
Skip
|
| Back to Top |
|
| |
dhoward
Admin Group
Joined: June 29 2001
Location: United States
Online Status: Offline
Posts: 4447
|
| Posted: April 04 2015 at 19:57 | IP Logged
|
|
|
Skip,
Ive been getting the send mail error message as well.
Not every time but at least half of my posts get the
message. Something that I will have to look into with
the forum software.
With PowerHome shutdown, open the pwrhome.ini file and
go to the [System] section. Change the LoadCtlrs and
LoadPlugins parms both to NO. Save and launch
PowerHome. Lets see if it hangs with both controllers
and plugins not installed. If it works, shutdown and
try changing to just loading controllers. Lets see if
we can narrow it down to controllers or plugins (or
other) being the problem.
Dave.
|
| Back to Top |
|
| |
MrGibbage
Super User
Joined: October 23 2006
Location: United States
Online Status: Offline
Posts: 513
|
| Posted: April 04 2015 at 20:36 | IP Logged
|
|
|
OK, we are starting to get somewhere. Setting both to
"NO" allowed it to work for 20 minutes with no problem.
No surprise there, actually. Setting Controllers to NO
and plugins to YES also allowed PH to work for quite a
while with no issues. I figures out how to disable the
one plugin that I had through PH Explorer-Setup-Plugins,
so no surprise there either. But when I set Plugins to NO
and Controllers to YES, I get a freeze very shortly after
launch. Specifically, I figured out it would freeze right
after the first zone status change. So, on the next
launch, right after launch I would go to Help-About and I
figured out that I could disable the Elk controller
before any zone changes happened. And now it is running
for about ten minutes now without a crash.
But I need the elk controller working! So, what's the
next step?
By the way, thanks for the help, Dave! 9:30 on a Saturday
night, no less!
__________________
Skip
|
| Back to Top |
|
| |
dhoward
Admin Group
Joined: June 29 2001
Location: United States
Online Status: Offline
Posts: 4447
|
| Posted: April 04 2015 at 22:41 | IP Logged
|
|
|
Skip,
So it appears that the Elk M1XEP is what is causing
the hang. I have the same controller in my setup and
have not had a problem although I do sometimes get the
lost communication message.
I'll do a comparison between 2.1.4 and 2.1.5c and see
what may have changed in the Elk controller so perhaps
can pinpoint what is going on.
I'll let you know what I find.
Dave.
|
| Back to Top |
|
| |
dhoward
Admin Group
Joined: June 29 2001
Location: United States
Online Status: Offline
Posts: 4447
|
| Posted: April 04 2015 at 23:21 | IP Logged
|
|
|
Skip,
Well after looking at the code, alot has changed. The
version of PowerBuilder changed, the Catalyst winsock
control version changed, and the code for the Elk M1
controller was pretty much retooled. Im not going to
be able to locate the issue that way.
Since it "hangs" and then comes back, perhaps we can
narrow down a little on your system. Can you have the
PowerHome Status screen running and then watch the
execution queue to see if the hang occurs based upon
what is in the queue and what shows on the realtime
log. See if we can see any patterns there.
Next, if you can time it right, try going to File-
>Debug Info and capturing the data to the clipboard
and emailing the results to me. If possible, try to
get captures just before and just after the hang. If
you're able to do this, it will yield some good
internal information.
Last, are sending any commands to the Elk (controller
commands) or doing anything else programmatically with
the Elk? You say the hang seems to happen during a
zone status. Can you give me more info on what is
occuring with your system when this happens?
Appreciate it.
Dave.
|
| Back to Top |
|
| |
MrGibbage
Super User
Joined: October 23 2006
Location: United States
Online Status: Offline
Posts: 513
|
| Posted: April 05 2015 at 06:55 | IP Logged
|
|
|
I think I am getting somewhere. I think it is related to
the ph_geturl1() formula. As I was testing, disabling
triggers and lines in macros and whatnot, it seems that any
time I try to execute that formula, it crashes.
So, I went to test the url I am sending, and it seems that
the website is doing fine, no issues there that I can
identify, but I am not going to rule out a possible problem
with the website. Strange that it was working fine on
2.1.4, but now has problems on 2.1.5c. Really strange.
What is your email address anyway?
__________________
Skip
|
| Back to Top |
|
| |
MrGibbage
Super User
Joined: October 23 2006
Location: United States
Online Status: Offline
Posts: 513
|
| Posted: April 05 2015 at 06:57 | IP Logged
|
|
|
By the way, is there a SQL command I can execute to get the
macro name of any macro that has at least one formula that
uses the ph_geturl1 command? I have a lot of macros that
call that command. I guess I should learn how to write
"subroutine" type macros...
__________________
Skip
|
| Back to Top |
|
| |
dhoward
Admin Group
Joined: June 29 2001
Location: United States
Online Status: Offline
Posts: 4447
|
| Posted: April 05 2015 at 10:08 | IP Logged
|
|
|
Skip,
Not sure why that would cause a problem as that
function didnt really change other than upgrading to
the latest version of the Catalyst winsock control.
Let me know what your testing shows though and I'll
see what I can do. There are several options with the
ph_geturl1 function so let me know the particular
option you're using.
Email address is my username at myx10.com.
Easiest way to find where you've used a particular
function is the "Database where used" report under the
reports menu. Just wildcard your search like this:
%ph_geturl1%
For the SQL way to do it:
Code:
select id from macrodetail where send_keys like
'%ph_geturl1%' |
|
|
Hope this helps,
Dave.
|
| Back to Top |
|
| |
MrGibbage
Super User
Joined: October 23 2006
Location: United States
Online Status: Offline
Posts: 513
|
| Posted: April 05 2015 at 10:53 | IP Logged
|
|
|
That search report is pretty cool! I have never seen that
before.
I usually use it like this
ph_geturl1("URL",1,20)
Although, I did see one time where I used a 3,20, and I
have no idea why. I should write more comments in my code.
I don't even remember what those options are for. So far,
it looks like everything is OK on my server. I don't see
any issues. I see in my logs other users using it, and it
seems fine when I send manual commands via my browser. I
can try and find out if any of them are on 2.1.5c and if
they are using the same command.
__________________
Skip
|
| Back to Top |
|
| |
nick7920
Senior Member
Joined: March 04 2008
Location: United States
Online Status: Offline
Posts: 193
|
| Posted: April 05 2015 at 13:03 | IP Logged
|
|
|
Skip,
I had problem with ph_geturl1 when I upgraded to 2.1.5 and I
think I have posted somewhere on the forum.
What I had to use instead was ph_getitcurl.
mine to work with PHCA (android app) example : (url is
https)
ph_getitcurl("{PHCA_SERVER}&garage_st={GARAGE_MAIN}",20)
(side note)
Also after upgrade make sure you have all the global
variable copied and working, I had a problem with variable
which had data length of more than 255 character.
Nick
|
| Back to Top |
|
| |
MrGibbage
Super User
Joined: October 23 2006
Location: United States
Online Status: Offline
Posts: 513
|
| Posted: April 05 2015 at 16:30 | IP Logged
|
|
|
OK, I will give that a shot. Thanks, Nick!
__________________
Skip
|
| Back to Top |
|
| |
MrGibbage
Super User
Joined: October 23 2006
Location: United States
Online Status: Offline
Posts: 513
|
| Posted: April 07 2015 at 18:18 | IP Logged
|
|
|
ph_getitcurl fixed it perfectly! Game on!
__________________
Skip
|
| Back to Top |
|
| |
dhoward
Admin Group
Joined: June 29 2001
Location: United States
Online Status: Offline
Posts: 4447
|
| Posted: April 07 2015 at 22:08 | IP Logged
|
|
|
Skip and Nick,
Just wondering if you can try the ph_geturl1 function
with the different option types to see if any of those
options will work or not. Not really sure why only the
ph_getitcurl function works and it would help me to
pin down what is unique about it if one of the other
options on ph_geturl1 will also work.
Appreciate it.
Dave.
|
| Back to Top |
|
| |
MrGibbage
Super User
Joined: October 23 2006
Location: United States
Online Status: Offline
Posts: 513
|
| Posted: April 08 2015 at 08:10 | IP Logged
|
|
|
I just did some testing. Here's the command that I sent:
ph_geturl1( "{PHCA_SERVER}?username={PHCA_USER}&password={PHCA_PASS}&ala rmstate=" + ph_htmlescape("{ELK_ALARM_STATE_AREA1}"),0,20)
And this is how it was sent after the variable substitution
ph_geturl1( "https://phca-web.appspot.com/send?username=skip&password=XX XXXX&alarmstate=" + ph_htmlescape("No Alarm Active"),0,20)
I tried it for options
0,20
1,20
2,20
3,20
Option 0 failed somewhat gracefully without locking up, but it popped up a dialog box with the following error message:
An error occurred while processing the execution queue. Resetting execution queue and trying again.
*** Error Details ***
Error Number: 39
Object Name: uo_socketblob
Class: uo_socketblob
Routine Name: f_geturl
Line: 51
Text: Error accessing external object property remoteport at line 51 in function f_geturl of object uo_socketblob.
Options 1 and 2 both locked up the system for exactly 4 minutes each with no other information.
Option 3, however, was successful!
Edited by MrGibbage - April 08 2015 at 08:11
__________________
Skip
|
| Back to Top |
|
| |
dhoward
Admin Group
Joined: June 29 2001
Location: United States
Online Status: Offline
Posts: 4447
|
| Posted: April 09 2015 at 23:00 | IP Logged
|
|
|
Skip,
Really appreciate the testing. This info along with
data from another user has helped me to pinpoint the
issue with the ph_geturl1 function.
Basically, any websites requiring SSL are going to
have an issue with options 0, 1, or 2. If the website
is non-SSL then everything should work. Option 3 does
work for both SSL and non-SSL but will be slow for SSL
sites. If retrieving data from an SSL site, you should
set the timeout value to as small a value as will
work.
I'll have this fixed in the next version.
Thanks,
Dave.
|
| Back to Top |
|
| |
Active Topics 
Memberlist 
Search 
Help
Register 
Login
PowerHome Messageboard : PowerHome General
Topic: 2.1.5c upgrade not working for me
