Our director node has displayed this error message this morning.
Then at the update log I´ve seen that it looks to be some problem with certificates:
[2012-04-23 08:08:27.647 +02:00] SSL handshake with host: https://wpm.webwasher.com failed (error:00000000:lib(0):func(0):reason(0))!
[2012-04-23 08:08:27.647 +02:00] Check your update certificates under /etc/pki/tls/certs
[2012-04-23 08:08:28.544 +02:00] SSL handshake with host: https://wpm3.webwasher.com successful!
What should I do??
Thanks in advance.
Since the handshake with wmp3 worked, you are still getting updates, right?
Is this a permanent or a temporary error?
Do all of your appliances have this error?
You could try to update the certificates with yum update mwg-certs.
To investigate further, you could also check the differences of the certificates in /etc/pki/tls/certs beween working and non-working appliances.
Yes, since the handshake with wmp3 worked, we are getting updates.
This error just happened once on 2012-04-23 and only in our active director node.
Since that moment it haven´t ocurred again till this saturday and today.
This time the same error it´s been displayed but now in two our appliances (Active director node d pasive director node)
We´ve checked the differences of the certificates in /etc/pki/tls/certs between working and non-working appliances and they look exactly equals.
Any idea why is this happening?
Thanks very much.
have you already filed an SR with Support?
Can you maybe have a look into the log files (update, coordinator) and let us know if there is anything suspicious?
the discussion is almost a year old. I would be pretty surprised if our update servers would not serve updates for a year now :-)
So I assume it is some networking/connectivity issue. Can you share some more details from the update/core/coordinator logs? Do you have a Service Request with support to have a deeper look into the issue?
Your updates servers do server updates but fail at this address. It then tried another address and will be sucessful.
In turn, this makes the Performance all Yellow blocks instead of nice Green blocks because they are updating.
No network changes here, this just started all of a sudden.
i see this behavior alos pretty often at my proxies in the office and at my customers.
it is important to find out how often the alert is shown. I will try to explain how the process usually works, maybe we can find out some more details why the alerts occur.
When MWG speaks to wpm.webwasher.com it connects to one of three servers. wpm.webwasher.com has three IP addresses and - depending on which IP is returned via DNS - MWG talks to this IP. In this communication the server at wpm.webwasher.com returns a real server address which points to a specific server.
When MWG is unable to talk to the IP it obtained when looking up wpm.webwasher.com, it will throw an error. This may happen from time to time because one of servers behind the "wpm cluster" is restarted due to maintenance, maybe the provider hosting the server has a problem, etc. Especially in the last two or three weeks there were problems with one of the servers causing it to become unavailable temporarily.
When you see the issue on a daily basis for multiple weeks I could imagine that MWG cannot talk to one of the IP addresses returned for wpm.webwasher.com. When MWG has a problem to connect with the next attempt it tries a different IP and succeeds. wpm-webwasher.com resolves to multiple IP addresses:
18.104.22.168, 22.214.171.124, 126.96.36.199
A simple test may be to connect to MWG via SSH and try to connect to those IP addresses via wget and verify there is some response. Maybe it is just one of the hosts which is unavailable to you.
If you are seeing alerts every now and then it should be fine, although it may become annoying. As mentioned in the last two or three weeks there were some issues with one of the servers, which would explain why there are alerts on a daily basis, but at the moment it should be pretty calm.
I have been getting these about every two days for weeks and weeks. Had one this morning at 3am.
Hope they get this fixed soon. Thanks Andre, we appreciate you man.