Re: external system monitor to verify all services up

Not exactly plug n play, but workable.
I've seen the api & downloaded a sample pgm after the first incident.
We are lights out from 6pm to 7am, plus all weekend. Even in lights out, we
take a large volume of transactions, both in and outbound, plus users and
customers online.
In last 3 weeks have had 2 incidents - one where TCP failed to start, and
last night where TCP failed to end (stuck in NFS). We do a large amount of
processing from midnight to 7am, so any issue is bad. We have monitors for
QSYSOPR messages, that wake someone up, but neither of these incidents
issued a message...
Another thought is that we give the job xx minutes to complete, and issue
alert if not complete - but if smtp down, then no email going out the door.

Got your last email - we are Notes (for now, but it's going out (parent
corp decision)). Still, the concept can work outside of Notes.

Jim


On Mon, Mar 23, 2015 at 3:05 PM, <rob@xxxxxxxxx> wrote:

Most of these services monitor a specific port as shown in
NETSTAT *CNN
On the NETSTAT *CNN if you do "F14=Display port numbers" you can see the
numbers.
Too bad NETSTAT *CNN doesn't have an *OUTFILE option. As an alternative
you could try using this API:
List Network Connections (QtocLstNetCnn)

http://www-01.ibm.com/support/knowledgecenter/ssw_ibm_i_72/apis/qtoclstnetcnn.htm?lang=en-us
To list them out to some file.
Then a remote client could attempt a remote command, to generate the list,
followed by a download, and then look for ports you are interested in. If
the ports not listed then the port is down.
If you had multiple lpars you could monitor each other. You could even go
so far as the sockets programs by Scott Klement to test the various ports
for listeners.

IBM has a product, called EDGE. We use that as a smart router. For
example if you go to our intranet site of dekkohome.corp.dekko.com it
looks for port 80 serviced by one IP address on one lpar. If that port is
down then it goes to port 80 on another lP address served up by another
lpar and so on. You can either do it this way or load balancing between
all active port 80's on a list of IP addresses. I've not played with it
to see if it has notifications if a certain port on a certain IP address
is down.

You can also google port testing software and you'll get a bunch of hits,
including some free downloads.



Rob Berendt
--
IBM Certified System Administrator - IBM i 6.1
Group Dekko
Dept 1600
Mail to: 2505 Dekko Drive
Garrett, IN 46738
Ship to: Dock 108
6928N 400E
Kendallville, IN 46755
http://www.dekko.com





From: Jim Franz <franz9000@xxxxxxxxx>
To: Midrange Systems Technical Discussion <midrange-l@xxxxxxxxxxxx>
Date: 03/23/2015 02:23 PM
Subject: external system monitor to verify all services up
Sent by: "MIDRANGE-L" <midrange-l-bounces@xxxxxxxxxxxx>



Looking for an external system monitor to verify all the services up:
telnet
ftp
http
lpd
smtp
snmp
etc

It would need to be external so if IP is down, an alert can be generated.
Does the HMC have this capability? Needs to be able to send text alert or
email.

We have a 7 hour (midnight to 7am) incident where the ENDTCPSVR command
was
stuck on a NFS issue, so part of system still up, but part was down.

Jim Franz
--
This is the Midrange Systems Technical Discussion (MIDRANGE-L) mailing
list
To post a message email: MIDRANGE-L@xxxxxxxxxxxx
To subscribe, unsubscribe, or change list options,
visit: http://lists.midrange.com/mailman/listinfo/midrange-l
or email: MIDRANGE-L-request@xxxxxxxxxxxx
Before posting, please take a moment to review the archives
at http://archive.midrange.com/midrange-l.



--
This is the Midrange Systems Technical Discussion (MIDRANGE-L) mailing list
To post a message email: MIDRANGE-L@xxxxxxxxxxxx
To subscribe, unsubscribe, or change list options,
visit: http://lists.midrange.com/mailman/listinfo/midrange-l
or email: MIDRANGE-L-request@xxxxxxxxxxxx
Before posting, please take a moment to review the archives
at http://archive.midrange.com/midrange-l.