Glad you got something up and running. If you're willing to keep debugging the issue you're seeing with OpsCenter 3.2.1, can you try a browser other than Chrome?
In an effort to consolidate free help offered for our products we have decided to move these forums to a more widely used forum. Please use one of the following queries (or any combination):
- Cassandra: tag search or plain text search
- DataStax Enterprise: tag search or plain text search
- DataStax OpsCenter: tag search or plain text search
Opscenter 3.2.1 install is unuseable where 3.1.1 runs fine.(29 posts) (3 voices)
Sure, It appears I have the same issue with Firefox 23 in both normal and private browsing as well as a fresh install of opera 15. These are all on a windows 8 pro machine including a fresh vmware build. However, I tested with Firefox on a Xbuntu VM and the page loads correctly. I previously did not test with firefox on Xbuntu only windows 8 machines. Also Safari 6.0.5 renders correctly. Could this be an OS issue? I will spin up a XP VM and report back.
I also have same issue with firefox 18 on XP-x86 SP3.
This is looking a lot like an issue that i had with EC2 a month ago although that was limited in scope to Firefox.
Mike, do you have Nick Bailey's contact info? He should have submitted a bug report on or about 2 Aug 13.
I am aware of that bug report, but didn't realize it was related to this. I actually tried to personally reproduce the issue on a Windows 8 VM with FF 22, and was not able to.
Is 10.50.10.51 the host of the VM that the browser is running in, or is it the host/another VM? Not being able to access the /tcp route would cause the push mechanism between opscenterd and the browser to not work, and would certainly account for data not displaying properly. I would suspect the issue is networking/proxy related, rather than browser related.
It may or may not be related, however I did want make sure you were aware of it.
10.50.10.51 is opscenter/cassandra
10.50.10.52 is cassandra
10.50.10.53 is hadoop
10.50.10.54 is solr
these are all headless installs on a vmware cluster.
My windows workstation that does not work is a physical machine at 10.50.1.183, inside of that workstation is a VMware player guest running Xbuntu with a bridged connection. That VM guest works fine. All machines are on the same subnet with no firewall, proxy or ACLs of any kind being applied. I am going to pull up wireshark and do a comparison of working and non-working clients. Do you have other suggested next steps?
So if I'm understanding correctly, the UI works fine for any browser inside one of the other VMs, but not from the host machine? From the host machine, are you able to load the Performance section and see graph data? (that doesn't rely on push)
No, you do not understand correctly, it looks like VM dont matter. The UI works fine from any of the Linux machines inside of the VM. The UI works fine on any Linux or OSX machine outside of the VM. The UI does not work on windows inside the VM. The UI does not work on windows outside of the VM. Even running a linux VM inside of a windows machine results in a broken windows UI and a working linux UI. Same switchport, different results depending on OS.
Have the same issue.
In my case opscenter is behind nginx proxy, not sure if it is a reason.
I can describe my case if you ask.
Rolled back to 3.2.0.
3.2.1 and 3.2.2 don't work.
@eagle: We did have another customer that had an issue with nginx. They found that turning proxy_buffering off solved the problem for them.
@MJ: I don't see any mention of nginx from you. Perhaps an accidental omission? Or maybe you have some other network setting that's doing something similar with buffering?
I have no nginx in my environment.
Any thoughts about where to go from here?
MJ, we made a change in 3.2 in the way the UI communicates with opscenterd for real time data (we switched from long polling to streaming over a long lived XHR connection). I'm thinking that even though you're not using nginx, that change is what is causing you problems. We can verify via manually patching (if you haven't upgraded to OpsCenter 3.2.2, do that first).
In /usr/share/opscenter/content/js/ripcord/ripcord.js (or ./content/js/ripcord/ripcord.js), change the beginning of line 308 from: