3 Replies Latest reply on Mar 30, 2019 2:56 AM by Marie Johnson

    Email engine issue - 9.1.04

    Debasish Hazarika

      Hello All,


      Looking for some suggestions or if anyone has seen this before. I have a customer who just went live with a 9.1.04 (patch 002) ITSM environment. The application performance has been great so no issues with that (knock on wood)


      The primary concern is the Email engine. They are a Solaris shop and have a significant load of outgoing emails every day. The email engine works great until it hits a wall with this error when the email engine pauses for about 10 minutes and then starts processing emails again. This happens like every 20 minutes and this causes the backlog of outgoing messages to increase significantly


      DEBUG SMTP: exception reading response, THROW:
      java.net.SocketException: Connection timed out (Read failed)


      Although it appears to be a SMTP connection issue between the app server and the mail server, but I have been with the customer (and their infra team) pretty much the whole day and we haven’t been able to prove this theory. When this error occurs we have tried doing telnets to the mail server from the app server and were able to send out emails using SMTP commands.

      I feel this is an issue sending out certain emails. I have tried setting the SMTP timeout parameters in the Email daemon but that hasn’t helped. I am hoping there is a setting somewhere which will tell the email engine that even if you are not able to send out this particular email, set the Send Message to “Error” and move on to the next one without taking a pause.

      The customer was on 8.1 until last week and they haven’t seen this happen before. Their email volume hasn’t changed from 8.1 to 9.1



        • 1. Re: Email engine issue - 9.1.04
          Uwe Ryczek



          i've never seen this error before..

          ...but with the newer ARS versions there is a Service Failover mechanism and if you have more than one AR Server (aka Server Group) you can split the load (if you have different Incoming and Outgoing mailboxes) to different AR Servers.


          You only have to change the ranking in the form "AR System Service Failover Ranking"


          good luck ;-)



          2 of 2 people found this helpful
          • 2. Re: Email engine issue - 9.1.04
            Swanand Dharmadhikari

            Have you tried checking with network team since this seems to be a problem or latency with Remedy Email to Exchange communication.

            Since this is being the Solaris environment which also has the local send mail service.

            • 3. Re: Email engine issue - 9.1.04
              Marie Johnson

              So a “read“ time-out has nothing to do with  sending or connecting to the mail server.  The “read” means your remedy system cannot read What it needs to in order to complete the transaction for the java plugin.


              typically this is due to a lack of resources And/or limitations.  Version 9 versus version 8 does use more java memory, was that tuned correctly?  Do you have enough threads?  Have you ran thread logs and looked at that?   In the java lib directory there is a logging.properties file can you look at the level of logging you’re doing?  Default is typically set to severe so you won’t get much detail.  here is documentation on the logging Debugging options for the BMC Remedy Email Engine - Documentation for Remedy Action Request System 9.1 - BMC Documentati…


              Another point is if your plugin is crashing and armonitor is restarting it, it may not have all the environment variables set correctly like when remedy starts up.  Validate the CLASSPATH, JRE_HOME  and PATH are correct and that each java  process in armonitor has the correct java version and CLASSPATH.


              there were features in the old java versions which were sunset and in java 8 there are completely new parameters  to use to fine tune the performance , were those set?

              1 of 1 people found this helpful