ChannelAdvisor Tracking Updates Broken

[4444...@gmail.com]

In version 8 ChannelAdvisor tracking updates cause Order Manager to freeze to the point where the only option is to force quit the program. We reported this issue in May and Julia confirmed the issue.  The only solution has been to use version 7.4 for tracking updates only since those updates still work in 7.4. So now we are running 8 on all of our systems except 1 which is running 7.4 to send tracking updates. When then licensing server goes offline and we can no longer use 7.4, we are going to have major problems.  I have been conversing with Erich in ticket 19873 who just stated you were stumped and didn't know what it could be. Could someone please help with this situation as it is a major problem that will cause major issues for us if it is not fixed. I either need a fix for the problem or a work around that will work once the licensing server goes offline.

[Moe]

yes we have been experiencing the same issues since we upgraded to 8.012.
This is extremely concerning and I'd like to know if this was resolved in the latest release?

[Moe]

and yes, we were told to do the same by stone edge and have 1 workstation running on version 7 to send tracking updates.

[4444...@gmail.com]

8.013 did NOT fix the issue despite the changelog saying it did. Have not tried 8.014 but it does not mention this issue at all.

[Moe]

 It’s disappointing to hear that 8.013 did not resolve the issue as we were about to upgrade.
With the holiday season fast approaching, we need a fix as soon as possible. We need to hear back with a solution or workaround.  

[John Frazar]

Hi Moe and Anonymous,

I want to avoid a finger-pointing scenario between Stone Edge and Channel Advisor, but I believe it’s important to highlight that this issue appears to be on Channel Advisor’s side, not with Stone Edge Order Manager (SEOM). Here’s why:

We were able to reproduce the problem using Postman, entirely removing SEOM from the equation, and we still saw the issue occur. This strongly indicates that the root of the problem lies with Channel Advisor, as the same failures are happening independently of SEOM. We’ve sent Channel Advisor a video showing these failed posts, and they seem to be responding and taking a deeper look at the problem now.

The freezing behavior you’re experiencing in SEOM happens because the code has stopped while it waits for a response from Channel Advisor that may never come. To mitigate the symptoms, we can try a couple of workarounds: adjusting SEOM to timeout faster when no response is received, or slowing down the posting rate to avoid overwhelming Channel Advisor’s server.  It's possible that when posts are sent in rapid succession their servers are dedecting the DDoS response which leads them to not respond at all to the post... that's a pretty common security tactic.

That said, I believe the best approach is to address the problem at its source rather than relying on workarounds. Channel Advisor has started investigating further, and we should give them the opportunity to fix this properly. In the meantime, we’ll continue monitoring the situation closely.

Thank you for your patience and understanding as we work through this. Let’s keep the lines of communication open, and I’ll keep you updated on any progress

[Moe]

Thank you for the follow-up and for making sure this gets resolved as quickly as possible. I appreciate the detailed explanation, but I’m a bit confused as tracking numbers have always been sent without issues from Version 7, including from our Version 7 workstation. This suggests that it may not be entirely on ChannelAdvisor’s end.

That said, I’m glad to hear that ChannelAdvisor is looking into it further, and I hope the issue can be resolved promptly. 

[4444...@gmail.com]

Thank you for the reply and for taking a deeper look into the issue.  I agree with Moe that If this is a ChannelAdvisor issue then why does sending tracking updates work in 7.040?  Did the timeout settings change from 7.040 to  version 8? 

[Paul Christel]

Do those with issues process large numbers of CA orders?   Because there was no need, I never updated our Shared Stack OM machine to v8 along with the rest of our machines.  Your posts reminded me that I needed to do so by yesterday (thank you).  I updated our Shared Stack machine to 8.005 yesterday and haven’t had any issues with inventory adjustments or tracking number postings.   We’re not a high volume store and am wondering if we’ll see the issue pop up when we get more busy in Q4.

 

Paul

--
You received this message because you are subscribed to the Google Groups "Stone Edge User Group" group.
To unsubscribe from this group and stop receiving emails from it, send an email to stoneedge+...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/stoneedge/d7153291-1806-41a5-9b3f-070865ee01f0n%40googlegroups.com.

[4444...@gmail.com]

We process 50-130 CA orders daily. We've tested with every number combination of order and the number of orders doesn't seem to matter. We've had it freeze on a single order or batches of orders. Sometimes it will send 10 through without issue and then freeze on a single order. We've tried using the setting that sends tracking after each shipment, so one at a time, and that also causes it to freeze. Inventory adjustments don't seem to be a problem. It is only tracking updates that have the issue. It doesn't freeze on every single order, but if you're sending more than a handful a day it is very likely to freeze on you.

[Paul Christel]

We send the tracking number after each order is packed and didn’t have any problems again today.  We probably had about 25 CA orders.

 

We had some weird connectivity issues earlier in the year, which we believe was caused by Comcasts security edge.   We now have it disabled, which resolved the issue.  This was mostly an issue on importing orders from CA, however.

 

However, I have no idea why it would work on an older version of the software.  Are you using the shared stack?

To view this discussion on the web visit https://groups.google.com/d/msgid/stoneedge/23ee367c-3e5c-41a2-8846-bddb13a101bfn%40googlegroups.com.

[Moe]

We have tried both shared and not shared. We are having the exact same issue as "anonymous". SE Support had already looked at it, acknowledged the problem and suggested using version 7 - which is what we are doing.
This needs to be resolved ASAP unless we can still use version 7 beyond the deadline given by Stone Edge.

[Moe]

I should add that we are using Version 7 for 1 workstation only just to send out tracking.
All others are on v8.

[Paul Christel]

Is it all ship methods or only specific methods/carriers that cause issues?

 

 

Paul

 

From: ston...@googlegroups.com <ston...@googlegroups.com> On Behalf Of 4444...@gmail.com

Sent: Tuesday, October 1, 2024 3:46 PM

To view this discussion on the web visit https://groups.google.com/d/msgid/stoneedge/23ee367c-3e5c-41a2-8846-bddb13a101bfn%40googlegroups.com.

[4444...@gmail.com]

The shipping methods are the same between 7.040 and 8.013, so shipping methods should not be the issue since 7.040 works properly. Since John confirmed there is an issue, I guess we will just have to wait for a solution at this point. 

[Paul Christel]

The methods should be the same since those are stored in the server db, but there is a local table (ShippersMethods) that stores the ship method codes sent to CA.  I find myself having to update this table almost everytime we upgrade.   When there is an issue with the values in this table, it usually doesn’t hang as you describe, but it won’t update the tracking in CA.  If the ones that work on occasion are always the same method, maybe that will help isolate the problem.

 

If you can get your hands on 8.005, I can confirm it’s worked without issue for two days now.  We ship USPS and UPS.  That would at least keep you from having trouble w/ the license server.

 

@John Frazar – Did the stack controller code change between v8.005 and 8.012?

To view this discussion on the web visit https://groups.google.com/d/msgid/stoneedge/e4a8e212-01c6-4497-89dd-7e32df739cben%40googlegroups.com.

[4444...@gmail.com]

That's interesting that 8.005 is working for you. We have only tried 8.008 and above and it hasn't worked in every version. Thanks for the info.

@John Frazar - Could we get a link to download 8.005 to test? At least then we'd have a working solution after the licensing server goes offline.

On the issue with the ShippersMethods table, that was partially fixed in one of the later releases. They are still putting the wrong name for Ground Advantage Cubic.  They are putting "USPS Ground Advantage Cubic" instead of "Ground Advantage Cubic". So if you do not change it to the correct name then the system won't send any tracking updates for that service to ChannelAdvisor. They seem to have fixed regular Ground Advantage in the update however.

That issue isn't related to the freezing issue however. For that issue it will just not send the tracking update to ChannelAdvisor at all if the names are wrong.

[Kevin Horowski]

I just wanted to add a quick update for this.  We are still talking with Channel Advisor support on this issue.  Stone Edge uses built-in VBA and Windows functions to communicate with the web so that limits what we can access when troubleshooting.  We were able to reproduce the issue outside of Stone Edge though by sending requests through Postman.  After each request, we usually get a success response just like what Stone Edge would be expecting.  However; after multiple posts we sometimes don't get back a response at all.  It's completely random.  I can send the same post 50 times within a minute and it will work fine. Then if I try the same post 5 more times on the 5th attempt we might get the issue with no response.  We thought maybe accounts were errantly being throttled due to the amount of updates being sent in a given period but Channel Advisor doesn't think it's any type of throttling as they said there should be a response in that situation with an error message/number that gives more info.

We did have one modification to the Stack Controller code between versions but it was related to the form opening and nothing with the sending of data.  Without looking back at the changes I believe we removed that code for unrelated reasons in version 8.013.

@4444...@gmail.com I will contact you directly with instructions on how you can try 8.005 if you'd like to see if you have the issue in that version.

To view this discussion on the web visit https://groups.google.com/d/msgid/stoneedge/SA1PR16MB585266F43E9DC7413664992CBE702%40SA1PR16MB5852.namprd16.prod.outlook.com.

[4444...@gmail.com]

8.005 does NOT work for us. It froze on the very first tracking number we sent.  We've had to go back to 7.040 for the time being. 

[Moe]

Just wanted to check to see if there are any updates on a resolution to this issue.

[steve]

I have several Stone Edge customers not having any issues with their shared stack and they run thousands of orders a week through the stack sending updates to Channel Advisor.  Even though this might sound counter to what's been reported, it's not broken everywhere.  Maybe trying to figure out why some people aren't having any issues might be another avenue to go.

[Moe]

Thank you for sharing your input. I can only say that everything has been working fine with Version 7 on the stack. We are currently on 8.012 but had to revert back to Version 7 on the stack workstation because tracking updates weren’t being sent. This issue was acknowledged by Stone Edge at the time, and they suggested using Version 7 as a workaround, mentioning that other clients had reported similar problems.

My concern now is that with the announcement regarding Version 7 being unsupported after December 31st, we may run into issues if a permanent fix isn’t implemented by then. It’s clear that this is impacting others as well, since “Anonymous” is facing the same issue.

I hope the team can find a long-term solution before the deadline, as relying on this older version is not a viable option moving forward.

[4444...@gmail.com]

For those of you who it works for, what API URL are you using for the CA shopping cart? We are using:

webadapters.channeladvisor.com/Export/CFExport.aspx?

Also, what version of Office are you using? We are using Office 2021 professional. 

I'm wondering if the issue isn't Office version related. I'm a but perplexed however as StoneEdge has been able to replicate the problem outside of Order Manager so it doesn't make any sense why it works for some and not others. 

We really need to get this resolved.

[John Frazar]

We have been continuously back and forth with CA on the issue.  Here's what we know:

1. The failures are completely random.  When it does fail, we can retry the same post a second later and it will work. 
2. Most SE/CA users do not have the issue
3. For those who do have the issue, it seems to work fine in some versions of SEOM but not others, despite the code being identical. 
4. We can recreate the issue outside of SEOM. 

#4 is the smoking gun that makes me think we should be looking at things outside of SEOM to troubleshoot, so I'll add to your questions:

For those of you who it's working or not working, can you please post information about your operating environment?

• What Operating System do you use?
• Are you on WiFi or Ethernet?
• Any firewall/malware/anti-virus software or hardware on your PC or LAN?
• Which ISP do you use?  

To view this discussion on the web visit https://groups.google.com/d/msgid/stoneedge/7d5533ff-958f-4ebc-9e38-6d9c6571e79an%40googlegroups.com.

[4444...@gmail.com]

Thank you for an update.

 Our settings where it does not work are below:

Operating System: Windows 10 Pro 

Wired Ethernet

No Firewalls at all on PC or router

ISP Wide Open West (WOW)

Office 2021

If it was a firewall, ISP, or OS issue then it shouldn't work in version 7 on the same machine, but it does.  The bizarre part is that Version 7.040 is working fine, so that tends to suggest that there is some incompatibly between the OS or Office version and Version 8. 

[Paul Christel]

Our cart (v8.005) is simply set-up with: webadapters.channeladvisor.com.

 

Our stack controller is running Office Version 2019 Pro.

To view this discussion on the web visit https://groups.google.com/d/msgid/stoneedge/7d5533ff-958f-4ebc-9e38-6d9c6571e79an%40googlegroups.com.

[Paul Christel]

I tried changing the API URL to see what happened on my end, but it seems to me that the server actually used is hardcoded to api.channeladvisor.com and not using the cart setting.

To view this discussion on the web visit https://groups.google.com/d/msgid/stoneedge/SA1PR16MB5852DBD45154EDED25529982BE792%40SA1PR16MB5852.namprd16.prod.outlook.com.

[4444...@gmail.com]

Have there been any updates to this issue?

[Paul Christel]

Did you try temporarily disabling any AV software on the stack controller machine?

To view this discussion on the web visit https://groups.google.com/d/msgid/stoneedge/14e89292-fcb4-4227-bcd2-f917ae1565a9n%40googlegroups.com.

[4444...@gmail.com]

Yes, there is no Antivirus software or any type of firewall running on the computer or router. It works in 7.040 on the same computer so that eliminates firewall or antivirus also.

[bradr...@gmail.com]

Hello. It's troubling that so many others are experiencing issues.

We upgraded to 8.011 and haven't noticed any issues, or very, very few. I have seen the import machine freeze randomly while processing tracking, but very infrequently.

Windows 10 Pro. Comcast is our ISP. Ethernet

[Moe]

V8.012 - stack does not send to CA for us
V7.022 - stack works

Operating System: Windows 10 Pro 

Wired Ethernet

No Firewalls

Verizon fios

Office 365

[4444...@gmail.com]

Any update on this? If not can some type of work around be built into OM so we can use the program when the licensing server goes offline?

[4444...@gmail.com]

Could we please get an update on this ChannelAdvisor issue. We need to have a solution in place before the licensing server goes offline!

[Jaclinne]

We r also waiting for solution for sales tax import not import correctly tooo and we cannot upgrade yet and we already spend 1500 move our customization to the original version - and now facing software server about to close down 🥹 during our peak season 

(Stone edge monthly pay user since 2005) 

-Jaclinne 

On Nov 18, 2024, at 8:42 AM, 4444...@gmail.com <4444...@gmail.com> wrote:

Could we please get an update on this ChannelAdvisor issue. We need to have a solution in place before the licensing server goes offline!

To view this discussion visit https://groups.google.com/d/msgid/stoneedge/c4c4aa23-43a2-4a07-bc3c-8870bb29ccb4n%40googlegroups.com.

[4444...@gmail.com]

Any update on the solution for this issue? The last I checked a solution was being worked on. We're coming up to the point where the licensing server is going offline so we'd like to have a fix in place ASAP before that happens. 

[Kevin Horowski]

We are still working on a new code update which should hopefully be tested by the end of the day today.  We still aren't sure what is causing the issue.  We've already updated the Channel Advisor code for Tracking Updates to their latest version.  The current version in Stone Edge is deprecated but still does work for some people for some reason.  Even after the update to the latest version of their integration the issue where no response is being received by Channel Advisor was still occurring.  The newest code change that we are finishing is related to the way we send the updates.  Most updates in Stone Edge for communication to channels, marketplaces and payment gateways go through HTML post which uses a built-in Windows DLL called wininet.  We are updating the code for Channel Advisor orders to go through the WinHTTP service.  The updates will still go through the stack controller for monitoring.

To view this discussion visit https://groups.google.com/d/msgid/stoneedge/5d819ddb-825d-4726-9b02-f98363f0ab6bn%40googlegroups.com.

[Moe]

  Thank you for the update. Will this be released as a patch to our current version, or will it be part of a yet-to-be-released OM version update?
We are currently on version 8.012 and were planning to upgrade to 8.014, but we will hold off if a new release will include this CA update.  

[Kevin Horowski]

The code changes are small enough and are limited to just a few places in the program and isolated to Channel Advisor so we could treat it as a patch.  As of now, the code changes are in the ChannelAdvisor Functions module and the Stack Controller form.  I don't think most people would have modifications in either of these places so it would just be a matter of replacing the module and the form but if someone just wants to replace the new code we can do that too.  We will probably just reach out to everyone we know with the issue and offer to make the changes.  If anyone would want to apply the changes themselves we can also make them available.

To view this discussion visit https://groups.google.com/d/msgid/stoneedge/b49e253d-e173-4def-86b8-c64adbd6e4ebn%40googlegroups.com.

[Moe]

Just an update - We upgraded to version 8.014 and put in those CA code changes and ............so far so good.
Trackings seem to be updating on CA but we are still currently monitoring just in case. Thanks for the fix Kevin.

[4444...@gmail.com]

Is this update still working for you? While it is "working" for us, we are routinely getting On Timer errors now after processing tracking numbers which cause us to close and reopen Order Manager.  

[Moe]

Interestingly, we experienced an issue earlier today where the tracking numbers weren’t sending. We tried restarting the program, but it still wouldn’t send. Eventually, we restarted the machine, reopened Order Manager, and it finally went through.

This was the only instance we’ve encountered with this issue since updating the code.

Not sure if both our issues are related as it happened to me today...perhaps CA was down for a bit?

[4444...@gmail.com]

Thanks for the info. The code definitely does not work for us. The OnTimer errors happen regularly and seem to have replaced the spinning wheel unfortunately. Could you let me know what version of SQL you are running? That's about the only thing I can think of that may be different. Thanks.

[Moe]

I cant say that its been working flawlessly. Just today we found a bunch of tracking updates that were left on the stack since yesterday.   We had to reopen OM and hit send now and it went through. This happened several times already since the code update. I believe we are using SQL 2016 lite version.

[Moe]

Spoke to soon. We are also getting OnTimer errors. We got that error yesterday and again today.
Seems like we are experiencing the same issue. Hoping that SEOM can resolve this.

[4444...@gmail.com]

Sorry you are having the On Timer issues too. StoneEdge stopped replying to my emails about a week ago and has been ZERO help.  It appears the new code they provided is having the exact same issue as before, but where we used to get the spinning wheel we are now getting the On Timer error. I believe the problem is that occasionally the StackController is timing out on the http.send command for ChannelAdvisor. The new code adds a timeout period for the POST and when it hits that timeout it is throwing the On Timer error.  I believe the new code needs some error handling now to say that if it hits this timeout retry the http.send because right now it just errors out.   So if StoneEdge reads this, please add timeout handling to the code!

[4444...@gmail.com]

I may have found a temporary solution. I have not thoroughly tested this yet so be wary.  In the new StackController code, before ".http Open" and after ".http Send" put "On Error Resume Next" . This will cause it to just put an error in the stack when it fails WITHOUT crashing the program.  The error in the stack will have the Channeladvisor URL as the last response. It should then keep trying to send the tracking updates until it sends out.  I'll post again after I've had a chance to thoroughly test it, but that won't be until next week. Good luck.

[4444...@gmail.com]

Just to update...Adding the code above to the new StackController completely fixes the issue. I've tested extensively now and the On Timer errors have not reappeared and tracking numbers ARE properly updated.  Also, still no replies to any of my emails to StoneEdge in about 2 weeks.

[steve]

Bravo!

[steve]

I implemented the fix above for one of my customers and it's been working well since the change.

[Paul Christel]

Has anyone figured out the common thread regarding customers experiencing the issue vs. those that haven’t seen it at all?

 

Paul

To view this discussion visit https://groups.google.com/d/msgid/stoneedge/dac9b1cd-b96c-4ee8-b073-d28cd8d34e77n%40googlegroups.com.