update check out of sync with running version

[ananke]

This is our first upgrade, so I'm not sure if it's normal or not; however, we repeated it in both testing & production. After upgrading to 1.4.2, main page shows a warning about new version being available, despite the fact that we're running that version.

[hsmith]

Are you by chance behind a proxy? I cannot reproduce this issue.

[ananke]

Are you by chance behind a proxy? I cannot reproduce this issue.

No proxy, direct access to the outside world. I do however notice that one of the system shows the correct 'You're running the latest [...]' after being rebooted, which would imply something didn't get restarted. Here's the procedure that was used on both of them:

Code: Select all

# upgrade
cd /tmp
rm -rf nagioslogserver nagioslogserver*.tar.gz
wget http://assets.nagios.com/downloads/nagios-log-server/nagioslogserver-latest.tar.gz


# Disable shard allocation ¿ this only needs to be run on a single node:
curl -XPUT localhost:9200/_cluster/settings -d '
{
 "transient" : {
 "cluster.routing.allocation.enable" : "none"
 }
}'

# unpack installer
tar xzf nagioslogserver-latest.tar.gz
cd nagioslogserver/
./upgrade


# if it goes ok, sleep for 10 seconds
sleep 10


# re-enable sharding
curl -XPUT localhost:9200/_cluster/settings -d '
{
 "transient" : {
 "cluster.routing.allocation.enable" : "all"
 }
}'

[hsmith]

This might have to do with a small issue regarding the browser cache. If you run into anything else clear your cache and let us know.

Is your issue good to go for now?

[ananke]

This might have to do with a small issue regarding the browser cache. If you run into anything else clear your cache and let us know.

Unfortunately, it's not a browser cache issue. I can see the same mismatch after logging in from a different system on a different browser. All signs point to something needed to be restarted, the question is which component? Steps listed above follow the upgrade procedure from:

https://assets.nagios.com/downloads/nag ... Server.pdf

[hsmith]

Is this the same system as your other thread?

[ananke]

Is this the same system as your other thread?

No. This issue we observed on two NLS systems we've had running for a few weeks. Other issue is on a third system (we're deploying it until eventual feature will exist to allow privilege separation for log content).

I could try to restart individual pieces of NLS to see which one triggers proper refresh; however, as mentioned earlier, I've stuck to the documented upgrade procedure. I figured one time is a fluke, but encountering it on two systems means there may be an issue.

[hsmith]

I think we found what is causing this, and it should be fixed in the next minor bugfix release. Is everything else working okay otherwise?

[ananke]

I think we found what is causing this, and it should be fixed in the next minor bugfix release. Is everything else working okay otherwise?

Thanks! So far so good, for other issues I'm creating new posts.

Cheers!

[tmcdonald]

In that case I'll be closing this thread, and we can continue in the others.