I have a php cli shell script that I want to be running continuously without ever stopping. However, I noticed that it would stop executing after a while on its own.
Is there a way to keep a script running forever without timing out, or daemonize it?
i have had a problem for some time now, regarding my CRON jobs. I am trying to download a large amount of data from ebay (through their API, totally legal and aboveboard) using php, but my CRONjob times out.
I have tried resetting the timeout variable, but then it exceeds the maximum filesize
SO, my question: is there any way to have a script run as a CRON job, and wen it is complete, call another script?
I'm not sure why my brand new dual proc quad core xeon 2.5ghz harpertown gets time out when the server load is under .5
Like it'll be running ultra fast and suddenly, I can't get into ssh, whm, my websites or anything. When I ping it, no reponse. Is it because it restarted itself?
I have recently tweaked my server (AMD 3000+ 1gb Ram 10mbps port) by configuring the httpd.conf, my.cnf and php.ini.
I am pleased to say that the server is now responding well and the load is always below 1.00.
However, sometimes a user will experience a time-out through their browser. Once they refresh the server will then react as it should and carry out the command which is being asked of it.
tweaking apache so that timeouts do not occur....here are the changes I have made...
php.ini
Code:
[MySQL]
; Allow or prevent persistent links.
mysql.allow_persistent = Off
httpd.conf
Code:
KeepAlive On
MaxKeepAliveRequests 100
KeepAliveTimeout 3
MinSpareServers 16
MaxSpareServers 32
StartServers 20
MaxClients 150
MaxRequestsPerChild 5
and finally here is my.cnf file
Code:
[mysqld]
skip-locking
skip-innodb
max_connections = 400
key_buffer = 200M
myisam_sort_buffer_size = 64M
join_buffer=2M
read_buffer_size = 5M
sort_buffer_size = 5M
read_rnd_buffer_size = 5M
table_cache = 1536
thread_cache_size = 128
interactive_timeout=100
wait_timeout=10
connect_timeout=10
tmp_table_size = 48M
max_allowed_packet = 16M
max_connect_errors = 10
query_cache_limit = 3M
query_cache_size = 64M
query_cache_type = 1
thread_concurrency=4
log-slow-queries = /var/log/mysql/mysql-slow.log
old-passwords = 1
[mysqld_safe]
open_files_limit = 8192
[mysqldump]
quick
max_allowed_packet = 16M
[myisamchk]
key_buffer = 64M
sort_buffer = 64M
read_buffer = 16M
write_buffer = 16M
I decided to merge 2 old dedicated servers into 1 colocated machine with better specs the old machines have a combined number of about 280 accounts.
I purchased a Broadberry Server and requested a specific setup and OS (CentOS) after some delays they finally got it working and shipped it to the Datacentre
I chose 49pence/RapidSwitch for colocation in the UK
I received a email from 49pence on how they wanted it setup and Broadberry did this as well which was good.
Unfortunately I got the email it had been received and installed from RapidSwtich before I received the email regarding server admin and password info
Broadberry set a very weak password up a bit of an oversight this
As within 12 hours of it being installed it was hacked!
Being a UK bank holiday we were unable to do anything till today
And now we are having to retrieve the server to reinstall everything and start again!
I hope the companys involved will be cooperative so we can get this up and running asap.
My severs at Coreix end later this month.
A lesson to be learned for us and I hope anyone reading
Next time we have a check list to make sure nothing gets overlooked!
Luckly no data on the server and no harm done other than cost and time.
I had a strange error this morning, httpd was running fine but nothing was loading. All the other services worked fine but I checked the error log and couldn't find anything. I restarted httpd and it's running fine now.
Quote:
[Sat Feb 10 11:48:01 2007] [notice] suEXEC mechanism enabled (wrapper: /usr/sbin/suexec)
[Sat Feb 10 11:48:01 2007] [notice] Accept mutex: sysvsem (Default: sysvsem)
[Sat Feb 10 13:06:02 2007] [notice] caught SIGTERM, shutting down
[Sat Feb 10 13:06:03 2007] [notice] Apache configured -- resuming normal operations
[Sat Feb 10 13:06:03 2007] [notice] suEXEC mechanism enabled (wrapper: /usr/sbin/suexec)
[Sat Feb 10 13:06:03 2007] [notice] Accept mutex: sysvsem (Default: sysvsem)
[Sat Feb 10 20:42:26 2007] [notice] caught SIGTERM, shutting down
[Sat Feb 10 20:42:28 2007] [notice] Apache configured -- resuming normal operations
[Sat Feb 10 20:42:28 2007] [notice] suEXEC mechanism enabled (wrapper: /usr/sbin/suexec)
[Sat Feb 10 20:42:28 2007] [notice] Accept mutex: sysvsem (Default: sysvsem)
Looks just like normal operations... I checked the access log and nothing looked out of the ordinary either.
Anyway the only suspicious thing I saw was the daily scan by spammers to see if I had anything exploitable.
Quote:
[Sat Feb 10 00:16:32 2007] [error] [client 69.13.76.82] File does not exist: /var/www/html/a1b2c3d4e5f6g7h8i9/nonexistentfile.php
[Sat Feb 10 00:16:33 2007] [error] [client 69.13.76.82] File does not exist: /var/www/html/adxmlrpc.php
[Sat Feb 10 00:16:33 2007] [error] [client 69.13.76.82] File does not exist: /var/www/html/adserver/adxmlrpc.php
[Sat Feb 10 00:16:33 2007] [error] [client 69.13.76.82] File does not exist: /var/www/html/phpAdsNew/adxmlrpc.php
[Sat Feb 10 00:16:33 2007] [error] [client 69.13.76.82] File does not exist: /var/www/html/phpadsnew/adxmlrpc.php
[Sat Feb 10 00:16:33 2007] [error] [client 69.13.76.82] File does not exist: /var/www/html/phpads/adxmlrpc.php
[Sat Feb 10 00:16:33 2007] [error] [client 69.13.76.82] File does not exist: /var/www/html/Ads/adxmlrpc.php
[Sat Feb 10 00:16:33 2007] [error] [client 69.13.76.82] File does not exist: /var/www/html/ads/adxmlrpc.php
[Sat Feb 10 00:16:33 2007] [error] [client 69.13.76.82] File does not exist: /var/www/html/xmlrpc.php
[Sat Feb 10 00:16:34 2007] [error] [client 69.13.76.82] File does not exist: /var/www/html/xmlrpc/xmlrpc.php
[Sat Feb 10 00:16:34 2007] [error] [client 69.13.76.82] File does not exist: /var/www/html/xmlsrv/xmlrpc.php
[Sat Feb 10 00:16:34 2007] [error] [client 69.13.76.82] File does not exist: /var/www/html/blog/xmlrpc.php
[Sat Feb 10 00:16:34 2007] [error] [client 69.13.76.82] File does not exist: /var/www/html/drupal/xmlrpc.php
[Sat Feb 10 00:16:34 2007] [error] [client 69.13.76.82] File does not exist: /var/www/html/community/xmlrpc.php
[Sat Feb 10 00:16:34 2007] [error] [client 69.13.76.82] File does not exist: /var/www/html/blogs/xmlrpc.php
[Sat Feb 10 00:16:34 2007] [error] [client 69.13.76.82] File does not exist: /var/www/html/blogs/xmlsrv/xmlrpc.php
[Sat Feb 10 00:16:34 2007] [error] [client 69.13.76.82] File does not exist: /var/www/html/blog/xmlsrv/xmlrpc.php
[Sat Feb 10 00:16:34 2007] [error] [client 69.13.76.82] File does not exist: /var/www/html/blogtest/xmlsrv/xmlrpc.php
[Sat Feb 10 00:16:35 2007] [error] [client 69.13.76.82] File does not exist: /var/www/html/b2/xmlsrv/xmlrpc.php
[Sat Feb 10 00:16:35 2007] [error] [client 69.13.76.82] File does not exist: /var/www/html/b2evo/xmlsrv/xmlrpc.php
[Sat Feb 10 00:16:35 2007] [error] [client 69.13.76.82] File does not exist: /var/www/html/wordpress/xmlrpc.php
[Sat Feb 10 00:16:35 2007] [error] [client 69.13.76.82] File does not exist: /var/www/html/phpgroupware/xmlrpc.php
I have nothing to be exploited so I'm thinking that wasn't the cause either.
I checked user_beancounters and there are also 0 fail counts.
Quote:
Version: 2.5
uid resource held maxheld barrier limit failcnt
509: kmemsize 4679572 9150217 50000000 50000000 0
lockedpages 0 0 256 256 0
privvmpages 51822 84271 262140 524280 0
shmpages 10913 10929 21504 21504 0
dummy 0 0 0 0 0
numproc 52 91 400 400 0
physpages 14273 24587 0 2147483647 0
vmguarpages 0 0 131070 2147483647 0
oomguarpages 14912 33040 26112 2147483647 0
numtcpsock 16 146 360 360 0
numflock 1 6 188 206 0
numpty 2 2 16 16 0
numsiginfo 0 25 256 256 0
tcpsndbuf 154284 371176 1720320 2703360 0
tcprcvbuf 197308 539932 1720320 2703360 0
othersockbuf 13416 239892 1126080 2097152 0
dgramrcvbuf 0 2800 262144 262144 0
numothersock 13 27 360 360 0
dcachesize 0 0 2273280 2416640 0
numfile 1199 2022 5820 5820 0
dummy 0 0 0 0 0
dummy 0 0 0 0 0
dummy 0 0 0 0 0
numiptent 89 89 128 128 0
I'm working on a script to help users get routed to the nearest, fastest server for the best ping. I'm in 2 datacenters, one on the east and one one the west coast of the US.
I've looked at some of the geo lookup programs based on IP, but they either seem inaccurate or expensive.. and just downright difficult to use.
I found out that some geo load balancers use the connection speed to figure out the best route. So, I'm trying to think of a way of timing the users connection from multiple server locations.
Has anyone here done that sort of thing before? Any suggestions on how to best do that?
Two completely different methods I've considered:
1. putting 2 images on a web page, and using javascript to time the loading of them.
2. pinging the user IP from each coast and seeing which is fastest. (Is there a lighter way than ping? )
There is a behavour with my server FTP when uploading a whole directory with many files in many sub-directories
Very often, the server disconnect itself when actively uploading files and the log simply says 'timeout'
It is as if the file got 'stuck' half way, and the FTP consider them as idle, therefore it disconnect you with a 'timeout' before reconnecting you.
But i have no problem uploading a single 200mb file to the server via FTP. I suppose no problem with 'keep alive'
So what is this behavior and how to solve it?
First time configuring Apache (2.2 w/ mod_jk) and when running locally (192.168.1.x), no problem.
When configuring from a DynDns through my UVerse router, I get several JS and CSS 'not found' and of course bad rendered site.
I have a VPS located in LA, USA.
For over a week now I have had the following network issues:
- browser timing out (for me and visitors to my site)
- ftp connection issues
The server load is low so it's not server related.
Traceroute TO the server appears fine.
Traceroute FROM the server to users IP's appears to have issues over the SingTel/Optus network.
My webhost says it's an issue for SingTel/Optus.
SingTel/Optus Engineer say:
"Our testings point to a problem either within Cogent's network or on a peering link between Cogent and Singtel in LA.
I'd suggest that the owner of the domain (me!) approach his hosting provider and have them escalate to Cogent. We can't escalate to Cogent as we have no peering with them."
So I've been the meat in the sandwich for over a week with no sign of a fix.
My options appear to be to either move the VPS away from the webhost and host it locally (Australia) or to somehow wait for someone to step up and take responsiblity and get this resolved.
My heart says wait as it's not *my* responsibility but it's costing me financially and professionally.
Anyone else experiencing similiar/same issues from the Asia Pacific region to the US?