Timing of Intial Sync

[SarahS]

I am working to implement MirrorSync to an existing solution.  The solution is currently being used locally on a MacBook.  The changes I am making will require the solution to be hosted in order that it can be accessed by several MacBooks, devices, as well as the internet.

 

I am working on the hosted files to develop the solution & implement the MirrorSync tool.  These files currently contain test data.  The current data still resides on the user's MacBook and she still needs to use the program while I am implementing the changes.  Will I run into problems by implementing & debugging MirrorSync with the test data, then later importing the current data and rerunning the sync tool?  If so, what procedure do you recommend?

[Jesse Barnum]

That's no problem - go ahead and developer with test data. When you're ready to use live data, use the MirrorSync admin utility to reset the sync data. To do this, right-click on the name of the configuration and select 'Reset SyncData' from the menu. This will invalidate all current offline files, but will start fresh with the real data for any new offline files.

[SarahS]

Wow, you made that simple, thank you!

[SarahS]

Was re-reading your response & wanted to verify the process.  When I am ready to implement the live data, do I need to import it into the server files & then sync it from the server to the offline files?  The data is currently stored on a MacBook in a different file structure (files that were never downloaded from MirrorSync).  I need clarification as to which files to import the live data to (if this is necessary), when to reset the SyncData, & when/where to initiate the inital sync from?  Thank you!

 

P.S.  If there is another forum posting that addresses this already & I didn't find, feel free to point me to that.

[Jesse Barnum]

1) Import the live data to the server

2) Reset the sync data

3) Distribute offline copies to users (I recommend using the download link feature of MirrorSync to accomplish this)

4) Run the sync script on the offline copies

[SarahS]

Hello Jesse,

 

This past weekend was the go-live with my solution in Mexico!  We used the tool successfully on Saturday (including synchronizing the data back to the server after the day was done) and there were some changes I needed to make for Sunday and ran into a few issues.  I have a couple questions & areas that I would appreciate having clarified.  I don't seem to see answers to these questions in the online documentation or existing forum postings.

 

1) What is the purpose of the "Disable link" button found within the dialog after selecting the "Download Database" button?  Is that for when an employee or person that previously had access to the files has left & you don't want them to have access anymore?  I have been disabling links & creating new ones as I make changes, but that doesn't seem necessary...

 

2) On Sunday, I downloaded new, revised files & installed them on the iPad, but was not able to do the initial sync.  I am confused as to when a device needs to be removed (in the licensing screen).  Do you need to remove a device & let it be readded during the initial sync once new files are installed?  Because I had to abort the initial sync (it seemed to stall out & I didn't have time to wait it out), I replaced the files with a fresh copy of the new ones I had just downloaded & we used them like that on the mini iPad.  We have not attempted the initial sync to the server yet b/c I didn't know if there would be issues due to using the solution prior to doing an initial sync.  Do we need to remove the device as one of the syncing devices prior to doing the initial sync?  Should we save a copy of the files to the hard drive as a backup before attempting to run the initial sync?  These files would not be device-specific yet since they have not been synced from the iPad, correct?

 

3) I was able to do conflict resolution on the MacBook, but not the mini iPad.  Should it work to do the conflict resolution on the mini iPad?

 

4) Is there a way to set a rule as to how you want conflicts resolved for one particular field?  I expect them to occur more often for that field and want it handled the same every time.

 

These are the questions I called about & was hoping to receive a call back about, so you don't need to call me if you understand my questions & can respond via the forum.

 

Thank you,

Sarah

[Jesse Barnum]

Hi Sarah - glad you e-mailed to remind me, I was watching the USA / Germany game and forgot all about calling you back - sorry!

 

#1: Yes, you would use the disable link feature if somebody who had the link was no longer authorized to download files anymore. By disabling all existing links and issuing a new link to only authorized users, that would remove the ability of somebody who had the link from using it. It's not necessary to do this every time you want to generate a new link, and in fact it could cause problems if other people (who you want to have access to download the files) are still using the old link.

 

#2: You never need to remove devices in the license panel unless you've reach the limit on the number of devices, and want to make a slot available for some other device. If the initial sync is not working, please file a bug report so that I can review the log file. It is OK to use the solution prior to doing the initial sync, although it simplifies things and makes the initial sync faster if you do the initial sync shortly after downloading the databases from the server.

 

#3: Yes, it should. I don't have an iPad mini to test with though, so if you're running into problems, please let me know more details.

 

#4: No, we do not support setting conflict resolution rules at the field level.

[SarahS]

Thank you for your response/answers.  They were very helpful!

[SarahS]

Hello Jesse,

 

The user in Mexico did the initial sync last evening (she hadn't been using it a lot since the go-live weekend) & ran into a couple problems.  At first there were some conflicts she had to manually resolve, and was able to do that successfully.  After completing the manual conflict resolution, she synchronized the iPad to the server, which seemed to go fine.  When she synchronized her laptop to the sever, however, all the sales from Sunday were synchronized to be in duplicate.  This was only true on the laptop.  The ID numbers have a two-digit device prefix and she informed me that the duplicate records each have a different prefix ID.  She wrote:

"I checked Saturday’s sales and they are fine in both the MacBook and the iPad. 

 

As for Sunday’s sales, the numbers are different. On the iPad all start with 08 except for one that starts with 05. On the MacBook, some start with 08 and some with 18."

 

I did a little more research this morning... When I opened the remote files on the server, I see Sunday's sales orders all have the prefix 05 (which is the correct prefix for records created on the server, but these records were created on the devices with the prefix 08 & 18), so I'm not sure what is going wrong... I have the ID Serial & ID fields both defined to have the "not empty" validation unchecked.  I would appreciate a little guidance and hope that you understand my description!

[Jesse Barnum]

First of all, let me mention that it's a good idea to do the initial sync as soon as the offline copy is generated. That will eliminate problems like this. If you call the 'MirrorSync Setup' script from your startup script, that will trigger the initial sync to happen whenever the offline file is opened for the first time.

 

As for fixing the problem, I need to get a copy of your server log file. This is in C:Program Files360WorksApplicationslogs on Windows or /Library/360Works/Applications/logs on Mac. If both syncs happened on the same day, then get the log file from that day - if they were on different days, then get both log files from both days.

 

Please upload the log files to http://360works.com/upload

[SarahS]

Thank you Jesse.  I didn't think there would be an issue after my question #2 above when I asked about doing the initial sync after making changes... I'm sorry to hear that it was an issue.  I uploaded two log files for you.  One from Sunday 6/22/14 which will perhaps reveal why the initial sync seemed to be stalling out after downloading the new files, and one from yesterday 7/2/14 when the initial sync was completed.  I had the user save a copy of the iPad files to a folder on her MacBook yesterday PRIOR TO doing the initial sync, so perhaps those files could be used to revert back to since they shouldn't be device-specific yet.  I don't know how it all works, so you can let me know if that is the best option (or even an option!)