Kinetic won't load - sits at white screen or spinning blue circle gets stuck

Kinetic ERP
122

Yesterday, Epicor support recommended that we clear all of the local path locations clearing the cache in this KB0035493 - Clearing the Client Cache in Kinetic/ERP

However, this morning a user who has been having this same issue daily since Sunday’s update had the issue happen again.

Overnight Epicor sent me another message saying in addition to clearing cache to try changing the client *.sysconfig setting for MaxBOMRU and MaxClssAttrMRU to zero in addition to stating that their dev and support teams are still investigating this issue and are just trying to provide workarounds right now. I haven’t tried changing the client *.sysconfig setting for MaxBOMRU and MaxClssAttrMRU to zero yet. Has anyone tried this and seen it fix the actual issue for good?

123

This just hit us today. Not sure why we were not affected by it earlier in the week. Lots of people needing the cache folders deleted.

3 Likes
124

It may not be the same, but it wouldn’t be the first time the same new bug was introduced across multiple versions. We didn’t have the issue with 2023.2.24 and earlier, just after we bumped to .29. I dropped a batch file on the most affected users desktops that clears all three cache locations so they could self-fix.
I’m going to look into the MaxBOMRU and MaxClssAttrMRU settings for us, but with the holidays I doubt I’ll get any data points for a while.

1 Like
125

Just adding what might be a titbit. 2 users running Epicor clients on their laptops were asked to update their passwords. 1 running windows 11 needs to have Epicor reinstalled. 2nd running windows 10 pro was able to log in after 15 min timeout for forgetting password…
The commonality being password expiration.

Just checked with 3rd user. Epicor could open… go figure.

126

We seem to attract all the bugs, so the fact that we’re not being hit by this issue is telling. I may have some information that could help.

Starting 3+ years ago, on the Epicor-hosted cloud, we would get random Business Layer Exceptions on all pages (including home), processes would get stuck and wouldn’t unstick, the perpetual login wheel daily…all of it. Support could not/would not figure it out, so I complained and complained and complained until I got more senior analysts looking at the problems…and what they discovered was the Collaborate connection was failing, and that in turn was hosing up communications everywhere. Once we completely stopped/removed Collaborate, most of the big issues went away. It had something to do with the Synergy service…I can’t quite remember, but it was a service that was (I believe) intended to keep authentication between different services in sync.

Anyway, my point is…check your logs and check Collaborate. This is the type of error we were seeing:
***<Op Utc="2024-09-03T11:49:28.9475490Z" act="Ice:LIB:Synergy/SynergySvcContract/GetAvailableGroups" correlationId="0c18e785-a93d-43a9-ad2b-ce32c2dc495e" dur="13.9924" cli="10.43.1.xxxx:1111" usr="jacob" machine="LVxxxxxx" pid="18428" tid="88">*** ***<Exception><![CDATA[Ice.Common.DataValidationException: Invalid SysConfig data. There is no SynergyId.***

Collaborate was updated on 12/15 - the same day/day after the 2024.2 upgrade. Just saying…

1 Like
127

We don’t have Collaborate. I would be pretty salty if a product we don’t have is affecting us

2 Likes
128

We also don’t have collaborate. We also don’t use basic auth so its not a password issue.

1 Like
129

I think he was referring to windows login, so SSO.

130

Well we use basic auth so we have both cases covered :laughing:

131

Not a fix but a workaround I have been using is having them sign in to an account that is working, for me its a generic shop floor account, once their signed into that they go to settings and click change user, type in their user info and it will load.

2 Likes
132

@Olga this is the workaround you had suggested, right?
image

No dice:

image
image1296×719 19.6 KB

2 Likes
133

I had one user that didn’t mention they were having problems. Apparently after 30 minutes of trying they were able to get in…

2 Likes
134

Why didn’t they ask for help! :scream:

135

His users like him more than yours like you?

4 Likes
136

Now that’s just mean.

2 Likes
137

I think you also have to set MaxBOMRU value to 0

1 Like
138

Tim writes above to set both values to 0
image
Maybe this is the important difference

1 Like
139

Why though. Moving that file makes no difference, its only the other one that has to be moved to fix the issue.

1 Like
140

Tim suggested two settings.

For the record, we just added the clear cache batch file I mentioned here to all our computers Startup folder and a shortcut to their desktop. We’ve not had any tickets about it since.

141

I understand, but those settings correlate to two different xml files in the cache. In our testing, moving the BOMRU file does not solve the problem. Only the class file has to be moved. So that would seem to indicate that only the class setting is relevant.

1 Like