Bug #13768 | Embedded MySQL Server is not re-entrant. (crashes) | ||
---|---|---|---|
Submitted: | 5 Oct 2005 10:49 | Modified: | 24 Nov 2006 13:45 |
Reporter: | Gary Ashford | Email Updates: | |
Status: | Verified | Impact on me: | |
Category: | MySQL Server: Embedded Library ( libmysqld ) | Severity: | S4 (Feature request) |
Version: | 5.0.13 rc | OS: | Windows (Win 2000) |
Assigned to: | CPU Architecture: | Any |
[5 Oct 2005 10:49]
Gary Ashford
[5 Oct 2005 11:06]
Vasily Kishkin
Thanks for test case. I was not able to reproduce the bug on 5.0.14 rc. Could you please verify your test case on 5.0.14 rc ?
[5 Oct 2005 13:43]
Vasily Kishkin
Sorry my mistake. I was able to reproduce the bug. I need to analyze the bug more.
[5 Oct 2005 14:40]
Heikki Tuuri
Hi! InnoDB is not re-entrant in the embedded server. Unfortunately, it is not even in the TODO to make it re-entrant. Why would you want to shut down the embedded server and start it again in the lifetime of the process? Regards, Heikki
[5 Oct 2005 15:52]
Gary Ashford
Our application instanciates the server process dynamically under the control of commands written in our 4GL. Since we have no control of how end-users write their apps, it is entirely probable that a user may intentionally (or accidentally) terminate the server process and then try to re-start it again. Whether the embedded server is re-entrant or not, it should certainly not crash when an attempt is made to re-start it! This is not a TODO, but more of a general stability issue. I hope you will agree.