Zimbra Server 502 Error

Discuss your pilot or production implementation with other Zimbra admins or our engineers.
Post Reply
stringers
Posts: 13
Joined: Mon Mar 19, 2018 4:55 am

Zimbra Server 502 Error

Post by stringers »

Hi all,

I am trying to fix a problem occurring with my Zimbra server, which is currently at version 8.7. Please bear with my explanation as the error is a bit confusing to me as it seems to manifest in a variety of ways.

In short, users are having trouble with the HTTP server. There are two main symptoms of their problem:
  • When they try to send an email it fails, with the email reappearing in the user's drafts and the server sends a message too their inbox informing them off the crash, a 502 error.
  • When they try to access the web interface for their mailbox, it is extremely slow or crashes, giving the same 502 error.
The detailed error message for the first symptom looks like (edited for brevity):

Code: Select all

Error Code: <html>
<meta http-equiv="Content-Type" content="text/html; charset=ISO-8859-1"/> <title>Error 502 Connection to Upstream is Refused</title> </head> <body> <h2>HTTP ERROR 502</h2> <p>Problem accessing ZCS upstream server.
	Cannot connect to the ZCS upstream server. Connection is refused.<br/>
    Possible reasons:
        <li>upstream server is unreachable</li>
        <li>upstream server is currently being upgraded</li>
        <li>upstream server is down</li>
    Please contact your ZCS administrator to fix the problem.
<i><small>Powered by Nginx-Zimbra://</small></i><br/>
Error Text: <html>
<head>
<meta http-equiv="Content-Type" content="text/html; charset=ISO-8859-1"/> <title>Error 502 Connection to Upstream is Refused</title> </head> <body> <h2>HTTP ERROR 502</h2> <p>Problem accessing ZCS upstream server.
	Cannot connect to the ZCS upstream server. Connection is refused.<br/>
    Possible reasons:
        <li>upstream server is unreachable</li>
        <li>upstream server is currently being upgraded</li>
        <li>upstream server is down</li>
    Please contact your ZCS administrator to fix the problem.
<i><small>Powered by Nginx-Zimbra://</small></i><br/>
The second case is the same (returned as an html response to the browser with the following addition at the top:
Message: Network error Error code: AjxException.NETWORK_ERROR Method: AjxRpc.prototype._invoke Details:Failed to execute 'send' on 'XMLHttpRequest': Failed to load 'https://postoffice.wmawater.com.au:7071 ... oOpRequest
There are a few other oddities in the logs which I can go into as required. One standout example is that I can produce the above mentioned errors accurately when I see the following messages in /opt/zimbra/log/nginx.log (whilst using the tail -f command on the server). Many (without measuring I would guess at least 100 individual lines detailing the same thing, although from a variety of users) of these errors appear together:
2018/03/19 13:34:05 [error] 24587#0: *373878 no live upstreams while connecting to upstream, client: 192.168.100.132, server: postoffice.wmawater.com.au.default, request: "POST /Microsoft-Server-ActiveSync?Cmd=Ping&User=podger@wmawater.com.au&DeviceId=3F5BDBE5DC664803B12E849B59DB5139&DeviceType=WindowsOutlook15 HTTP/1.1", host: "postoffice.wmawater.com.au"
I am aware that similar questions have been asked elsewhere on the forums but none of these threads appear to have received conclusive answers. Specifically I have looked at this one which has similar symptoms: viewtopic.php?t=61967

Any thoughts are appreciated and any further clarification needed please ask.
Post Reply