MOD_PHP limits???

[Willem Grooters]

Environment: VMS 7.3-2, Apache 1.3-1, MOD_PHP 1.2-1; Patches applied.

One of the pages the PHP script generates is not finished for some reason.
The code at that point is generated by a PHP script that is included in the executed script.
Another PHP script includes exactly the same code but finishes completely.
I have tried to add some comments just before the location where the script ends (echo ) but then the script stops even earlier.
I therefore have the idea that there is some limit, likely in MOD_PHP, that causes the script to abandon - without warning.

APACHE$WWW settings are (still) default.

Module (included in other code) is attached; Location of error is marked.I added the resulting source. More info can be obtained (willem(at)grootersnet(dot)nl

[Chris Barratt]

Hi Willem,

Not sure if it is useful to this problem or not, but there is a setting called "output_buffering" that you can set up in your php.ini which controls when stuff is written back from PHP to the client browser. In our case, I set it to "on" so that I could control the header output for a particular page - the default was to write http headers back to client straight away.

Probably not related to your problem, but thought I'd mention it just in case.

Another question about the problem...does the apache process which is doing the PHP bit complete and go back to it's normal state or does it go into a CPU loop ?

Cheers,
chris

p.s. It was hard to read the attachment because when it was downloaded most of the top bit came as a single line - s removed somewhere.

[Willem Grooters]

Lack of buffer size won't be the problem, I think. The HTML code returned is about 3K7 bytes and stops all of a sudden as you may have noticed, and I have seen files returned that are twice as big (according Apache's access_log).
I don't think of a CPU loop, in the code I don't see why that would happen. It might be some process limit. There are quite a lot of modules loaded by the script (of which the given script is one). I raised PGFLQUO of APACHE$WWW but that didn't make any difference. Could it be buffers somewhere inside PHP? Global pages?
I'll dig through the apache$nnnnn.log files, will let you know.

[Willem Grooters]

Apache$$%%%%.log shows:

$ Set NoOn
$ VERIFY = F$VERIFY(F$TRNLNM("SYLOGIN_VERIFY"))
[Fri Aug 12 22:15:22 2005] [warn] NameVirtualHost 192.168.0.11:443 has no VirtualHosts
[Fri Aug 12 22:15:22 2005] [warn] NameVirtualHost 192.168.0.12:443 has no VirtualHosts
%SYSTEM-F-ACCVIO, access violation, reason mask=00, virtual address=0000000000000000, PC=00000000000000
00, PS=0000001B

Improperly handled condition, image exit forced.
Signal arguments: Number = 0000000000000005
Name = 000000000000000C
0000000000010000
0000000000000000
0000000000000000
000000000000001B

Register dump:
R0 = 000000000000001C R1 = 000000000000001D R2 = 0000000000000000
R3 = 0000000000000000 R4 = 0000000000000000 R5 = 0000000000000000
R6 = 0000000000000000 R7 = 0000000000000000 R8 = 0000000000000009
R9 = 0000000001E3EDD8 R10 = 0000000000000006 R11 = 000000000206A000
R12 = 0000000000000000 R13 = 0000000000B8E8F8 R14 = 0000000000B8ECBC
R15 = 0000000000B8ECB8 R16 = 000000007AE0BBF8 R17 = 0000000000000000
R18 = 0000000000000003 R19 = 0000000000000001 R20 = 000000007FFF0000
R21 = 0000000000000002 R22 = 0000000000000000 R23 = 0000000000000000
R24 = 0000000037DF55EB R25 = FFFFFFFF817BC200 R26 = 0000000000000000
R27 = 0000000000000004 R28 = 0000000000000000 R29 = 0000000000000000
SP = 000000007AE0BC30 PC = 0000000000000000 PS = 300000000000001B
APACHE$WWW job terminated at 12-AUG-2005 22:28:44.37

I learned via alternate cahnnels that there are some issues concerning PHP and SWS 1.3; these seem to be solved in 2.0 but I cannot upgrade, so I'll wait to 2.1 and get on from there.