Discussion:
Tomcat Manager weirdness on a Google Compute instance
James H. H. Lampert
2018-09-06 17:32:31 UTC
Permalink
(cross-posted to the gce-discussion Google group)

Ladies and Gentlemen:

I just experienced something very odd.

It seems that this morning, when I tried to upload a WAR file to the
prototype instance for one of my instance groups, I kept repeatedly
hitting a brick wall: as soon as the upload passed the 2% mark, it would
abort, and I'd get a "Connection reset" error page.

After some troubleshooting, following the Sherlock Holmes axiom (once
you eliminate the impossible, whatever remains, however unlikely, must
be true), I discovered that somehow, the "max-file-size" and
"max-request-size" in manager/WEB-INF/web.xml had reverted from our
standard setting of 500M, to the factory setting of 50M.

And I found that the same had happened with the prototype instance for
another instance group.

Can anybody shed any light on this? Some sort of automatic update?

--
JHHL

---------------------------------------------------------------------
To unsubscribe, e-mail: users-***@tomcat.apache.org
For additional commands, e-mail: users-***@tomcat.apache.org
Mark Thomas
2018-09-06 17:41:03 UTC
Permalink
Post by James H. H. Lampert
(cross-posted to the gce-discussion Google group)
I just experienced something very odd.
It seems that this morning, when I tried to upload a WAR file to the
prototype instance for one of my instance groups, I kept repeatedly
hitting a brick wall: as soon as the upload passed the 2% mark, it would
abort, and I'd get a "Connection reset" error page.
After some troubleshooting, following the Sherlock Holmes axiom (once
you eliminate the impossible, whatever remains, however unlikely, must
be true), I discovered that somehow, the "max-file-size" and
"max-request-size" in manager/WEB-INF/web.xml had reverted from our
standard setting of 500M, to the factory setting of 50M.
And I found that the same had happened with the prototype instance for
another instance group.
Can anybody shed any light on this? Some sort of automatic update?
Definitely not from any sort of automatic update on the Tomcat side.

Mark

---------------------------------------------------------------------
To unsubscribe, e-mail: users-***@tomcat.apache.org
For additional commands, e-mail: users-***@tomcat.apache.org
Louis Zipes
2018-09-06 17:42:54 UTC
Permalink
I discovered that somehow, the "max-file-size" and
Post by James H. H. Lampert
"max-request-size" in manager/WEB-INF/web.xml had reverted from our
standard setting of 500M, to the factory setting of 50M.
Does the file show a modified date/time that could point to at least a timing for the action?

-----Original Message-----
From: Mark Thomas [mailto:***@apache.org]
Sent: Thursday, September 06, 2018 1:41 PM
To: Tomcat Users List
Subject: Re: Tomcat Manager weirdness on a Google Compute instance

- - - external message, proceed with caution - - -
Post by James H. H. Lampert
(cross-posted to the gce-discussion Google group)
I just experienced something very odd.
It seems that this morning, when I tried to upload a WAR file to the
prototype instance for one of my instance groups, I kept repeatedly
hitting a brick wall: as soon as the upload passed the 2% mark, it would
abort, and I'd get a "Connection reset" error page.
After some troubleshooting, following the Sherlock Holmes axiom (once
you eliminate the impossible, whatever remains, however unlikely, must
be true), I discovered that somehow, the "max-file-size" and
"max-request-size" in manager/WEB-INF/web.xml had reverted from our
standard setting of 500M, to the factory setting of 50M.
And I found that the same had happened with the prototype instance for
another instance group.
Can anybody shed any light on this? Some sort of automatic update?
Definitely not from any sort of automatic update on the Tomcat side.

Mark

---------------------------------------------------------------------
To unsubscribe, e-mail: users-***@tomcat.apache.org
For additional commands, e-mail: users-***@tomcat.apache.org

---------------------------------------
CONFIDENTIALITY NOTICE: This message is for intended addressee(s) only and may contain information that is confidential, proprietary or exempt from disclosure. If you are not the intended recipient, please contact the sender immediately. Unauthorized use or distribution is prohibited and may be unlawful.
B�KKKKKKKKKKKKKKKKKKKKKKKKKKKKKKKKKKKKKKKKKKKKKKKKKKKKKKKKKKKKKKKKKKKKCB��[��X��ܚX�KK[XZ[�\�\��][��X��ܚX�P�X�] �\X�K�ܙ�B��܈Y][ۘ[��[X[��K[
James H. H. Lampert
2018-09-06 18:06:22 UTC
Permalink
Post by Louis Zipes
Post by James H. H. Lampert
I discovered that somehow, the "max-file-size" and
"max-request-size" in manager/WEB-INF/web.xml had reverted from
our standard setting of 500M, to the factory setting of 50M.
Does the file show a modified date/time that could point to at least
a timing for the action?
When I got the question, I'd already changed two of the prototype
instances, but there was one left.

That was the newest of the prototype instances, created August 16th. And
the date stamp on the manager/WEB-INF/web.xml is August 24th.

Which is odd, given that this prototype instance was created by cloning
another prototype instance, and loading it with different webapp contexts.

--
JHHL


---------------------------------------------------------------------
To unsubscribe, e-mail: users-***@tomcat.apache.org
For additional commands, e-mail: users-***@tomcat.apache.org

Loading...