Hello David, and hello Lawrin,
Lawrin, David from Microsoft Support has been helping me with this issue: TrackingID#2201060060003079
I attached my last email I sent to him to bring you up to speed.
I have been collecting debug diagnostic dumps and uploading them ([1]link to upload location) for him to analyze. This is his contact:
David Dietz
Senior Support Escalation Engineer
Developer Web Apps
Office: +1 (980) 776-7170
Working hours: Monday - Thursday 10am-9pm Eastern
David, Lawrin from mariadb.org is asking in her email below "Is this error what causes Application Pool to crash, or is it something that you get along the crashes?" I am not sure how to answer that.
The only thing I know is that IIS stops responding to requests, and that my logs look like this when IIS stops responding:
I forgot to mention that I temporarily set IIS to recycle every 25 minutes, in the hopes that IIS stays up longer:
Please find in bold my response between Lawrin's lines below.
Thank you,
Nathaniel Courtens
805 640-8883
-------- Forwarded Message --------
Subject: [JIRA] (ODBC-349) speed issues and IIS crashing
Date: Thu, 20 Jan 2022 11:16:00 +0000 (UTC)
From: Lawrin Novitsky (Jira) [2]<jira@mariadb.org>
To: [3]courtens@poweryourpoint.com
[ [4]https://jira.mariadb.org/browse/ODBC-349?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=211487#comment-211487 ]
Lawrin Novitsky commented on ODBC-349:
--------------------------------------
Thank you for the report.
Is this error is what causes Application Pool to crash, or is it something that you get along the crashes?
We need something, that could help us to find what causes the crashes.
The first step naturally is to get ODBC trace. In past it was a problem - IIS did not let the driver manager write to it. Not sure if that is changed, but worth of trying.
N: I suppose I could activate ODBC tracing. Should I start collecting these when IIS is not responding, or all the time? It could quickly produce a crazy amount of very sensitive data.
Also, would I need to be using defined System DSN for the tracing tool to work?
2nd option is if you could provide us with (minimal) standalone script/DB infrastructure, so we could repeat your problem locally.N: This is a bit more complicated.
If that is not possible, then the next option is create special driver version, that would somehow provide us with some debug information. If ODBC trace(from step1) is impossible to get, than that means, that we can't write (the debug info) to the file, thus we will need to figure out what to do.
speed issues and IIS crashing
-----------------------------
Key: ODBC-349
URL: [5]https://jira.mariadb.org/browse/ODBC-349
Project: MariaDB Connector/ODBC
Issue Type: Bug
Components: General
Affects Versions: 3.1.15
Environment: Windows 10 Pro with MariaDB 10.6, Apache 2.4 (Apache is mostly used as proxy), IIS 10, PHP 8.1, Code on IIS is ASP Classic (VBS). Hosted by VritualBox, dedicate 6 Cores and 5029 MB of memory. Reporter: Nathaniel Courtens
Assignee: Lawrin Novitsky
Priority: Major
Labels: Compatibility, crash, performance
I have been struggling finding the right ODBC driver after this server meltdown. I ended up upgraded from MariaDB 10.1 to MariaDB 10.6. Before the upgrade IIS was fast and reliable. I used to connect ASP to MariaDB 10.1 in two ways: MySQL ODBC 5.1 Driver, and MariaDB ODBC 3.0 Driver. Connection string was:
{{"PROVIDER=MSDASQL;DRIVER=
{MySQL ODBC 5.1 Driver}
;port=3306;SERVER=localhost;UID=;PWD=;database=_;Option=3"}}
{{"PROVIDER=MSDASQL;DRIVER=
{MariaDB ODBC 3.0 Driver}
;port=3306;SERVER=localhost;UID=;PWD=;database=_;"}}
–
Now on MariaDB 10.6 I have tried myodbc8w.dll MySQL ODBC 8.00.27.00 Unicode Driver which is fast but it crashes IIS after 1 or 2 days (not just the Application Pool but IIS!) I also tried maodbc.dll MariaDB ODBC 3.1.15 but it is slow (it takes 4x longer) and it crashes the IIS Application Pool every 2 to 3 hours in.
Link to my [related post|[6]http://mariadb.com/kb/en/asp-c0000005-errors-with-mariadb-connectorodbc-3115-64bit/]
I am getting errors such as:
MariaDB EventID 100: Aborted connection _ to db: '' user: '' host: 'localhost' (Got an error reading communication packets)
–
This message was sent by Atlassian Jira
(v8.13.13#813013)
----------------------------------------------------------------------------------------
[1] https://support.microsoft.com
[2] jira@mariadb.org
[3] courtens@poweryourpoint.com
[4] https://jira.mariadb.org/browse/ODBC-349?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=211487#comment-211487
[5] https://jira.mariadb.org/browse/ODBC-349
[6] http://mariadb.com/kb/en/asp-c0000005-errors-with-mariadb-connectorodbc-3115-64bit/
Thank you for the report.
Is this error is what causes Application Pool to crash, or is it something that you get along the crashes?
We need something, that could help us to find what causes the crashes.
The first step naturally is to get ODBC trace. In past it was a problem - IIS did not let the driver manager write to it. Not sure if that is changed, but worth of trying.
2nd option is if you could provide us with (minimal) standalone script/DB infrastructure, so we could repeat your problem locally.
If that is not possible, then the next option is create special driver version, that would somehow provide us with some debug information. If ODBC trace(from step1) is impossible to get, than that means, that we can't write (the debug info) to the file, thus we will need to figure out what to do.