Well put Rusty.  The host simply isn't there anymore when you get "no route to host"

That "route" is one of two things.  One, an arp entry for a local subnet host, broadcasting a "arp whois" query, and gets an "arp tell" if that host responds.  Two, an actual layer 3 route, either default or specific that says "go to this gateway to get there".  It still arps for that gateway in a local subnet too ala #1, that starts the chain.

Your host moved, or isn't there at the moment.  Using nmap -sP sweeps your local subnet with pings, and shows you the hosts, resolves dns if available (even mdns).  If yours isn't there, it moved, and should show you.  Look at your arp table as well via "ip neighbor" command.  If you know its mac address, you can find it that way.  Perhaps if this occurs in your network, you should keep the macs, and/or use dns.  

Do you have something else in the network handing out dhcp addresses?  Too short dhcp leases?  Not sure why your hosts are getting other addresses, or just falling off the network.  Unless the network is busted, which happens, perhaps try a new router.

Side note, I've had some chaos with like situation lately...  Screwing with security appliances and building a lab of many vm's, I spun up a windoze server, made it a dc, and thought to put dhcp and the full suite of crap on to play with AD in this decade, also to let windows dhcp sort out hosts and provide dns/ldap/etc.  Then a power outage or two quickly found that's a chicken and egg problem when my vm host reboots and dhcp goes away.  I had to move dhcp back and forth a few times between my firewall and the AD server, which caused a lot of movement of hosts to different addresses around my network, thus much of the same problem you're seeing now with my automation things, and this attaching to that within my network.

Yours sounds a bit familiar, something seems to be forcing that movement too.  That's not normal.

-mb


On Wed, Oct 31, 2018 at 9:34 AM Carruth, Rusty <Rusty.Carruth@smartm.com> wrote:
You might consider initiating the rsync from the backup machines, and even possibly on a cron job basis.

Also, I would have been surprised if the uninstall/reinstall of sshd would have helped, as its clearly a network issue, not an ssh issue - at least the 'no route to host' (which is an unfortunate choice of wording - it really means something like "I get no answer to an attempt to contact this host.  Either the host isn't there, or there is no route to it".)


Rusty

-----Original Message-----
From: PLUG-discuss [mailto:plug-discuss-bounces@lists.phxlinux.org] On Behalf Of Joe Lowder
Sent: Tuesday, October 30, 2018 5:51 PM
To: Main PLUG discussion list
Subject: RE: How to fix 'no route to host' problem

Thanks Rusty, Michael, and all ...

But I think I need a different solution than
"if all else fails" ;)

I have more than a dozen devices on my network
including four "back-up" computers to which I have
been using rsync to copy all of my most important
files on a regular basis. rsync is still working
fine to one of those back-up computers, but it
recently stopped working to 3 of the others and
this is the mystery I need to solve.

---------------------------------------------------
PLUG-discuss mailing list - PLUG-discuss@lists.phxlinux.org
To subscribe, unsubscribe, or to change your mail settings:
https://lists.phxlinux.org/mailman/listinfo/plug-discuss