Ask Questions and Find Answers
Important:
Ask is now read-only. You can review any existing questions and answers, but not add anything new.
But - don't panic! While ask is no more, we've replaced it with discuss - the new Liferay Discussion Forum! Read more here here or just visit the site here:
discuss.liferay.com
RE: "Sign In" pop-up populated with home page after new virtual instance cr
Hi there,
In brief when I am attempt to log into the core instance (I think this is the term that is used but maybe primary as well?) by clicking on the "Sign In" link the pop up I get contains the home page of the instance instead of a login form. If I go to "c/portal/login" I get a full login page but on login I get no error and I am simply redirected to the home page again which shows the "sign in" link.
I can't find anything relevant in the log file. I thought it might be cloudflare so disabled that but still no luck. This all started after I added a 2nd instance. I can log into the 2nd instance just fine with its admin account.
Anyone got any idea where to look? I have tried adding the "ominadmin.user=" to the portal-ext.properties but it didn't make a difference. I have a browser which has a remember me enabled and on that browser I can still get the navigation and product menu on the core instance but I don't have to login. Hope I don't ge tlogged out of that. All the other sites on the primary instance appear to work for guest but I cannot log in to them either with the admin user.
I am using the commerce plugin too - not sure if that somehow causes any issues.
thanks
In brief when I am attempt to log into the core instance (I think this is the term that is used but maybe primary as well?) by clicking on the "Sign In" link the pop up I get contains the home page of the instance instead of a login form. If I go to "c/portal/login" I get a full login page but on login I get no error and I am simply redirected to the home page again which shows the "sign in" link.
I can't find anything relevant in the log file. I thought it might be cloudflare so disabled that but still no luck. This all started after I added a 2nd instance. I can log into the 2nd instance just fine with its admin account.
Anyone got any idea where to look? I have tried adding the "ominadmin.user=" to the portal-ext.properties but it didn't make a difference. I have a browser which has a remember me enabled and on that browser I can still get the navigation and product menu on the core instance but I don't have to login. Hope I don't ge tlogged out of that. All the other sites on the primary instance appear to work for guest but I cannot log in to them either with the admin user.
I am using the commerce plugin too - not sure if that somehow causes any issues.
thanks
This might be (very) far-fetched but it does sound like an issue we had in our project after clustering our instances too.
In our case users would also experience pop-ups being populated with wrong contents, though we saw web-content lists where we should not. If I remember correctly, we had errors in how our multicast was set up, which, in short, led to irrational behaviour. So my suggestion would be to double-check if both your instances are actually properly clustered.
In our case users would also experience pop-ups being populated with wrong contents, though we saw web-content lists where we should not. If I remember correctly, we had errors in how our multicast was set up, which, in short, led to irrational behaviour. So my suggestion would be to double-check if both your instances are actually properly clustered.
Alexander BrinckmannHi Alexander,
This might be (very) far-fetched but it does sound like an issue we had in our project after clustering our instances too.
In our case users would also experience pop-ups being populated with wrong contents, though we saw web-content lists where we should not. If I remember correctly, we had errors in how our multicast was set up, which, in short, led to irrational behaviour. So my suggestion would be to double-check if both your instances are actually properly clustered.
I'd be *very* interested in more details, in case they're still available somehow. So far, I always expected errors in Multicast to result in too-early-cache-retirement, e.g. underutilization of cache, but not in this behavior. It'd be a great extra-check to add to the Devops documentation and training, to catch this early and when it doesn't erode trust in the software to do the right thing.
When I get the time to dig deeper I will try to provide all information that I can gather, it should be available somewhere still. We had a couple of issues regarding clustering, asynchronous behaviour, etc, and I wanna make sure to not mix things up. The way Mark described his problem sounded too familiar though.
Stay tuned.
Stay tuned.

Okay, after some research I think I can re-tell how we got to our problem.
First off, it was actually not directly linked to clustering and multicast issues itself. This seems to have been a coincidental overlap in events.
But a little more context:
We were using a Liferay CE 7.0.6 portal clustered in two instances. We use customized web-content structures that make use of nested portlets, creating articles that can have various "side portlets" nested in them. Most of these are web-content displays.
The error I remembered seemed to have happened when more than one user was editing these nested portlets at the same time, for example two people adding one nested portlet each to expand the 'article'. This could lead to a StaleObjectException, while output in the front-end could show unexpected behaviour, like populating pop-ups with wrong content.
https://issues.liferay.com/browse/LPS-82954 touches this issue.
So while the phenotype of the error seemed related, the underlying issue doesn't seem to fit this thread's original situation.
First off, it was actually not directly linked to clustering and multicast issues itself. This seems to have been a coincidental overlap in events.
But a little more context:
We were using a Liferay CE 7.0.6 portal clustered in two instances. We use customized web-content structures that make use of nested portlets, creating articles that can have various "side portlets" nested in them. Most of these are web-content displays.
The error I remembered seemed to have happened when more than one user was editing these nested portlets at the same time, for example two people adding one nested portlet each to expand the 'article'. This could lead to a StaleObjectException, while output in the front-end could show unexpected behaviour, like populating pop-ups with wrong content.
https://issues.liferay.com/browse/LPS-82954 touches this issue.
So while the phenotype of the error seemed related, the underlying issue doesn't seem to fit this thread's original situation.
Alexander BrinckmannThis might be (very) far-fetched but it does sound like an issue we had in our project after clustering our instances too.
In our case users would also experience pop-ups being populated with wrong contents, though we saw web-content lists where we should not. If I remember correctly, we had errors in how our multicast was set up, which, in short, led to irrational behaviour. So my suggestion would be to double-check if both your instances are actually properly clustered.
Thanks for the suggestion. I will have to read up on the mulicast bit as this is a sinlge instance. The phenomenon does seem to have someting to do with caching because I managed to login and get the admin panel once using Chromium, instead of firefox. The pop-up was populated correctly but on subsequent chromium tests it showed the incorrect pop-up again and the "c/portal/login" just redirected to the home page again.
I cleared out temp and work folders and deleted the osgi state cache but no change. I am investigating whether the change to recaptcha is linked to this problem as now none of our forms, on submission, show the success page, no entries make it into the entries list and no email is received. We are simply redirected to the home page.
I have tried changing back to "simple captcha" but I now no longer get a simple captcha image but just the text "Text to identify". I cannot conceptualise how this would affect logins and pop-ups but I am willing to follow up any path no matter how weird and, at the end of the day, I need to get the forms working again anyway

This is like the twilight zone.
It appears that the source of many of my problems was changing the default language for my instances from en_US to en_GB. This is reproducible see attached video.
I think this is the source of my problems as per post https://community.liferay.com/forums/-/message_boards/message/112820024
It has been identified as a problem for creating new commerce site per https://community.liferay.com/forums/-/message_boards/message/112820688
I think this is the source of my problems as per post https://community.liferay.com/forums/-/message_boards/message/112820024
It has been identified as a problem for creating new commerce site per https://community.liferay.com/forums/-/message_boards/message/112820688
Copyright © 2025 Liferay, Inc
• Privacy Policy
Powered by Liferay™