| Pages:
1
..
4
5
6
7
8
..
54 |
jptann
Member
 
Posts: 101
Registered: 13-6-2017
Member Is Offline
Mood: No Mood
|
|
Peter, in the next release are you going to fix the Custom Indicator portion? This blank editor started on 43.12
|
|
|
admin
Super Administrator
       
Posts: 5060
Registered: 7-4-2017
Member Is Offline
Mood: No Mood
|
|
Quote: Originally posted by jptann  | | Peter, in the next release are you going to fix the Custom Indicator portion? This blank editor started on 43.12 |
43.33 is pretty much ready for release, but it hasnt got this fixed. I think its going to be a simple fix, but we had much more critical bugs in the
new release. This is high on the todo list.
|
|
|
admin
Super Administrator
       
Posts: 5060
Registered: 7-4-2017
Member Is Offline
Mood: No Mood
|
|
GSB 43.35 build. March 1 2018
GSB Cloud fixes, support upload fix,
stability score in WF code, more diag info for errors
Custom indicator GUI fix.
Minor tweaks.
Includes settings for soyabeans
Includes a folder with gsbworker.cloud.exe if you want to contribute to GSB cloud.
dlls must be updated too from 43.25 and before
JPtann, this has your fix too.
|
|
|
jptann
Member
 
Posts: 101
Registered: 13-6-2017
Member Is Offline
Mood: No Mood
|
|
Thanks Peter, I will download them shortly.
|
|
|
jptann
Member
 
Posts: 101
Registered: 13-6-2017
Member Is Offline
Mood: No Mood
|
|
Peter, unfortunately the custom create a custom indicator has not been fixed. Your example for the MACD has 3 parameters for its function. To change
to my custom function, I have to go into the parameter area of the "generate custom" to delete the "close" variable and change the parameter range and
step for the next two variables. Unfortunately, your version 43.35 has screwed up in this area and does not allow me to change from 9:15:1 to
25:250:25. The older version of the code for this section worked fine, but something obviously simple changed here.
Thanks
|
|
|
jptann
Member
 
Posts: 101
Registered: 13-6-2017
Member Is Offline
Mood: No Mood
|
|
I have two parameters in my custom indicator that I have to change so i need this parameter settings to work like it used to.
|
|
|
admin
Super Administrator
       
Posts: 5060
Registered: 7-4-2017
Member Is Offline
Mood: No Mood
|
|
Quote: Originally posted by jptann  | | I have two parameters in my custom indicator that I have to change so i need this parameter settings to work like it used to. |
Sorry its not fully fixed.
can you send me a screen shot, as im not clear.
I removed the close and got this code in ts
NumToStr(MACD(9, 21) of Data1, 15) + "|" +
|
|
|
admin
Super Administrator
       
Posts: 5060
Registered: 7-4-2017
Member Is Offline
Mood: No Mood
|
|
I cant see the problem, but it might be fixed in 43.40 build. Hope to upload today.
|
|
|
admin
Super Administrator
       
Posts: 5060
Registered: 7-4-2017
Member Is Offline
Mood: No Mood
|
|
Please try this version, with installer. Give me feed back if your issue is fixed, and all else works well.
I'm planing to release this to trial users soon.
Attachment: Login to view the details
|
|
|
admin
Super Administrator
       
Posts: 5060
Registered: 7-4-2017
Member Is Offline
Mood: No Mood
|
|
client server docs updated. Probably should do a video on this sometime.
see http://www.trademaid.info/forum/viewthread.php?tid=17#pid184
|
|
|
cotila1
Junior Member

Posts: 78
Registered: 8-5-2017
Member Is Offline
Mood: No Mood
|
|
YES, please!! hugely apreciated
|
|
|
admin
Super Administrator
       
Posts: 5060
Registered: 7-4-2017
Member Is Offline
Mood: No Mood
|
|
GSB 44.03 build. March 7 2018
fix for some systems don't make it from the workers to the managers. I am not 100% certain this is fixed.
Has WF to the cloud, but manager & worker must be running (not stopped) for this to work.
For now you can only do 1 wf per active worker.
So if you want to wf 10 systems, you need 10 workers running.
Hope to make a job que and or multiple wf soon
|
|
|
Petzy
Junior Member

Posts: 73
Registered: 24-10-2017
Location: Sweden
Member Is Offline
Mood: No Mood
|
|
Great function to run wfo on workers!
One question. If I load an old system to my manager the choice to run wfo on a worker is greyed out. Is is supposed to be able to run loaded system on
workers as well?
|
|
|
admin
Super Administrator
       
Posts: 5060
Registered: 7-4-2017
Member Is Offline
Mood: No Mood
|
|
Quote: Originally posted by Petzy  | Great function to run wfo on workers!
One question. If I load an old system to my manager the choice to run wfo on a worker is greyed out. Is is supposed to be able to run loaded system on
workers as well?
|
The manager and worker have to be running to have the ability to send jobs to the workers. If the worker has stoped it gets disconnected from the
manager, and any manager can then again drive the worker. Im hoping for improvements to this in the near future.
Bug where workers systems dont make to the manager has been fixed.
|
|
|
admin
Super Administrator
       
Posts: 5060
Registered: 7-4-2017
Member Is Offline
Mood: No Mood
|
|
Ive noted that when manager finished, and no one else took the worker, WF jobs can still be sent to the worker.
WF from manager to worker didnt work with imported system.
There is a bug that the pf is missing from wf systems in the GUI with the graph. The pf is in the ts code however. Will be fixed soon.
|
|
|
cyrus68
Member
 
Posts: 171
Registered: 5-6-2017
Member Is Offline
Mood: No Mood
|
|
I'm not sure if I got it right regarding WF jobs. The requirements seem to be:
1. GSB must be set to operate in cloud mode
2. It must be actively generating systems
3. If you have 10 workers operating locally, then you can send 10 WF jobs to cloud workers
By the way 43.40 isn't that bad regarding missing workers.
yesterday I ran a model and out of 19,935 unique systems, only 6 didn't make it to the manager.
All the workers were local and active on one computer.
|
|
|
cyrus68
Member
 
Posts: 171
Registered: 5-6-2017
Member Is Offline
Mood: No Mood
|
|
sorry, I meant missing unique systems, not missing workers
|
|
|
admin
Super Administrator
       
Posts: 5060
Registered: 7-4-2017
Member Is Offline
Mood: No Mood
|
|
Quote: Originally posted by cyrus68  | I'm not sure if I got it right regarding WF jobs. The requirements seem to be:
1. GSB must be set to operate in cloud mode
2. It must be actively generating systems
3. If you have 10 workers operating locally, then you can send 10 WF jobs to cloud workers
By the way 43.40 isn't that bad regarding missing workers.
yesterday I ran a model and out of 19,935 unique systems, only 6 didn't make it to the manager.
All the workers were local and active on one computer. |
My failure rate of systems going missing was much higher. Its perfect now when I tested it.
What works right now, and what works soon are two different things.
I am not yet sure how much is implemented.
If you has only one worker, you should be able to send any amount of systems to the worker(s).
I sent 5 WF, but got two working, and 3 status is running. I'm not sure if the running ones are working. I have about 50 workers so too hard too find
the WF jobs.
However if the amount of wf per worker, is greater than x, or the ram is too low,it should put them in a job que.
I expect the job que is not yet done.
Hope this helps. Its work in progress.
|
|
|
cyrus68
Member
 
Posts: 171
Registered: 5-6-2017
Member Is Offline
Mood: No Mood
|
|
I ran a model on 44.03 and it is much worse than the previous build.
There were 6 workers on 1 computer and were operating properly. However the manager froze at 16/50 restarts.
The workers were at 28/50 restarts with 18,000 unique systems and the manager at 16/50 with 11,000.
The manager unfroze temporarily and then promptly refroze. When 50/50 was completed the manager was stuck at 45/50.
I need to go back to 43.40. However, I overwrote the GSB.pdb file with the new one. How do I get the old one?
|
|
|
admin
Super Administrator
       
Posts: 5060
Registered: 7-4-2017
Member Is Offline
Mood: No Mood
|
|
Quote: Originally posted by cyrus68  | I ran a model on 44.03 and it is much worse than the previous build.
There were 6 workers on 1 computer and were operating properly. However the manager froze at 16/50 restarts.
The workers were at 28/50 restarts with 18,000 unique systems and the manager at 16/50 with 11,000.
The manager unfroze temporarily and then promptly refroze. When 50/50 was completed the manager was stuck at 45/50.
I need to go back to 43.40. However, I overwrote the GSB.pdb file with the new one. How do I get the old one? |
You have way too many systems. You need to tighten up your performance metrics. I don't know if this is related but its the first thing I would look
at.
Going to the older build I expect will solve nothing.
You don't have to have .pdp file. The file does help however if you get a crash to diagnose things.
Can you zip up your exceptions folder of the last day or so and email to me. Last time you had issues it was connectivity to the sql server. No other
users have reported issues. Might be that again but dont know till I see the logs.
44.03 is going brilliantly for me.
see picture.
|
|
|
cyrus68
Member
 
Posts: 171
Registered: 5-6-2017
Member Is Offline
Mood: No Mood
|
|
the model was based on NG. the filters were the same as yours except that PF was reduced to 2.0.
I ran an NG model on version 43.40 with an additional secondary data stream and it worked fine.
the manager did not freeze.
I thought the role of the manager was to aggregate the systems produced by the workers.
speed was about 3500 per worker.
anyway i'll send you the files
|
|
|
admin
Super Administrator
       
Posts: 5060
Registered: 7-4-2017
Member Is Offline
Mood: No Mood
|
|
A web server that GSB cloud uses to get your IP address has fallen over. No idea when its going to come good.
The web site that failed is below. Cloud wont work and many
69.162.69.148:80
exception files will be generated.
Todays build fixes this.
|
|
|
jptann
Member
 
Posts: 101
Registered: 13-6-2017
Member Is Offline
Mood: No Mood
|
|
Peter:
ON the latest version of the system (stand-alone) 44.04, my computer last night did a reboot (always Microsoft on a Tuesday evening) and I had an
excellent running system but did not have time to review the WF completions. When I went to the auto save this morning, and tried to load a few of
the saved systems I got this error. How does one recover from such a disaster?
|
|
|
Gregorian
Junior Member

Posts: 97
Registered: 23-5-2017
Member Is Offline
Mood: No Mood
|
|
jptann:
You can successfully disable Windows Updates on Windows 10 with this tool:
http://www.intowindows.com/windows-update-minitool-alternati...
|
|
|
admin
Super Administrator
       
Posts: 5060
Registered: 7-4-2017
Member Is Offline
Mood: No Mood
|
|
Quote: Originally posted by jptann  | Peter:
ON the latest version of the system (stand-alone) 44.04, my computer last night did a reboot (always Microsoft on a Tuesday evening) and I had an
excellent running system but did not have time to review the WF completions. When I went to the auto save this morning, and tried to load a few of
the saved systems I got this error. How does one recover from such a disaster?
|
I don't understand this. Is your data and time correct?
Can you try to reboot. Can you try the manager. Windows update should not cause this.
If you have no success, zip up your c:\dropbox\gsb folder and put in dropbox or gdrive etc. If you send me an exe file it will be blocked by isp so
important to just send a url to GSB
|
|
|
| Pages:
1
..
4
5
6
7
8
..
54 |