On 10/12/2025 21:58, Linus Kamb wrote:
Mark,
Thanks very much for your response.
The disk is local, not shared.
So given the statement "... the disk from which all the tomcats run ..."
does this mean we are talking about multiple Tomcat instances running on
a single physical machine? Or something else?
Are any files shared between the instances?
How is the web application deployed? WAR file that Tomcat then expands?
I want to make sure I understand the architecture before I spend too
much time thinking about what might be going on.
Could the fact that there was no space left on the disk be treated as
being unavailable?
Seems unlikely. Tomcat has a list of files it checks for each webapp and
looks to see if the modification time has changed.
And why only the one webapp? Perhaps because it was the one attempting to
write to the disk?
Again, seems unlikely.
Might some external process changed the last modification time on any
files? A back-up process maybe? Those times should not be sensitive to
clock / time zone changes but I'd check if there were any changes of
that nature just in case.
Mark
- Linus
On Wed, Dec 10, 2025 at 1:07 PM Mark Thomas <[email protected]> wrote:
On 10/12/2025 18:52, Linus Kamb wrote:
All,
Context:
We have several public facing tomcats (all behind a firewall, no fronting
Apache.) They are mostly 10.1.44, but there is one 8.5.100. There is a
mix of webapps on all of them, but they all have one particular webapp in
common. That webapp often uses a fair amount of memory. (Not sure if
that
is relevant here.)
Situation:
The night before last *all* of the instances of that particular common
webapp were undeployed and redeployed in the middle of the night. (3 at
~1am, 1 at ~2am, and 2 at ~2:30.) There was no apparent nefarious
network
traffic, and in several cases, no network access to the tomcats at all
for
hours in either direction. And the timestamps of the war files and
web.xml
files had not (or do not appear to have) changed.
One possibly significant detail is that around that time, the disk from
which all the tomcats run was on the edge (and occasionally over that
edge)
of running out of space. For instance, in some cases where it didn't
already exist, tomcat reported it was unable to create the
catalina.2025-12-09.txt log file as there was no space on the device,
which
it was presumably trying to create to log the undeployment and
redeployment
of the webapp. (It was logged there in other tomcats where the file
already existed.)
There were no OutOfMemoryErrors thrown by any of the tomcats, and none of
the tomcats themselves were shut down and restarted. Just all the
instances of that one webapp in all the tomcats reloaded.
Any ideas what might have happened? Have you ever seen anything like
this? Can Tomcat spontaneously redeploy specific webapps unprompted?
Is there some form of shared disk arrangement between multiple machines
here? Is the web application on the shared disk?
If the shared disk goes off-line for any reason and auto-deployment is
enabled, Tomcat will undeploy the web application and redeploy it when
the shared disk returns.
Mark
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
