[BugDB] Runaway Children Memory Usage (PR#120)

1999-03-13 Thread bugdb-mod-ssl
Full_Name: Version: mod_ssl-2.2.4-1.3.4 OS: Linux2.2.2/Glibc2.1 Submission from: john.rightline.com (208.215.20.141) I have been succesfully running mod_ssl-2.0.13-1.3.3 and looking at upgrading to the current mod_ssl-2.2.4-1.3.4 and have discovered the following: 1) Memory usage for each

Re: ANNOUNCE: MM library, version 1.0b1

1999-03-13 Thread Jim Jagielski
Ralf S. Engelschall wrote: Mutex/Semaphore: o 4.2BSD-style flock() on temporary file o SVR4-style fcntl() on temporary file o SysV IPC semget() and send me your results. One thing I noticed that if all 3 mutexes are available, fcntl() will be chosen (since it's the last).

[BugDB] Runaway Children Memory Usage (PR#121)

1999-03-13 Thread bugdb-mod-ssl
Full_Name: Version: mod_ssl-2.2.4-1.3.4 OS: RedHat5.2 Submission from: john.rightline.com (208.215.20.141) PR#120 Just finished testing on RedHat5.2 standard install and have the same results as before. Triple the memory usage and child processes hanging.

Re: [BugDB] Runaway Children Memory Usage (PR#120)

1999-03-13 Thread Steffen Dettmer
Hi, Version: mod_ssl-2.2.4-1.3.4 OS: IRIX 5.3 here it looks good - no hanging processes, no high CPU comsume or so... Some seconds after a connect the load goes back to 0.01 or so... oki, Steffen 2) Child processes randomly hang and consume a large amount of CPU. The problem is that

Re: Network I/O errors w/ POST method

1999-03-13 Thread Ralf S. Engelschall
On Fri, Mar 12, 1999, Aaron Turner wrote: I saw the list archives that people are already aware of the bug, and was wondering what the situation was regarding a solution. I can generate the error on demand on my box using phpMyAdmin (a PHP3 MySQL management interface) and MOT (a PHP3 job

Re: [BugDB] Runaway Children Memory Usage (PR#120)

1999-03-13 Thread bugdb-mod-ssl
On Sat, Mar 13, 1999, Steffen Dettmer wrote: Version: mod_ssl-2.2.4-1.3.4 OS: IRIX 5.3 here it looks good - no hanging processes, no high CPU comsume or so... Some seconds after a connect the load goes back to 0.01 or so... 2) Child processes randomly hang and consume a large amount of