Server :: Logrotate For HTTPD Breaks HTTP Sessions
Dec 22, 2010
We started hosting some very large content on our site, and the usage patterns in cacti have revealed that the HTTP sessions through our load-balancers drop off dramatically right at midnight.
The logrotate process runs right at midnight, and issues a reload command through the service tool (CentOS 5.5):
Code:
$ cat /etc/logrotate.d/httpd
/data/websites/logs/*_log /var/log/httpd/*log {
missingok
daily
dateext
compress
rotate 7
sharedscripts
postrotate
/sbin/service httpd reload > /dev/null 2>/dev/null || true
endscript
}
Looking at the init script reveals that the reload section is suppose to trigger a HUP of the httpd process:
Code:
reload() {
echo -n $"Reloading $prog: "
if ! LANG=$HTTPD_LANG $httpd $OPTIONS -t >&/dev/null; then
RETVAL=$?
echo $"not reloading due to configuration syntax error"
failure $"not reloading $httpd due to configuration syntax error"
else
killproc -p ${pidfile} $httpd -HUP
RETVAL=$?
fi
echo
}
In which, Apache should reload it's configuration and start the new logfile without breaking current sessions. However that clearly isn't what is going on. I'm tempted to edit the logrotate script to trigger a HUP directly by cat'ing the PID file directly. Is this normal behavior for Apache when signaled with a HUP?
View 1 Replies
ADVERTISEMENT
Dec 10, 2010
I have an issue with my web server. We are running RedHat Enterprise Linux 3.0 with Apache 2.0 and Tomcat 5.5. The situation has arisen that the httpd sessions never terminate. New connections continue and continue to be created and never die. I have restarted the apache services to reset the connection and have even rebooted the server, however, to no avail. Yes, that does the trick of getting the web sites operational, however, this is not a solution.
I have searched and searched here, www.google.com/linux to no avail. I have looked through the apache.org bug tracker and can't find anything like what I am experiencing. This happened 6 months ago and I got lucky and it stopped, however, the situation has resurfaced. I have reviewed the logs and found nothing that provides any insight.
During the business day, the number of httpd connections continue to grow and I decided to let it see how high it would get before the web sites would crash. That magical number is 203. Now that it is later in the evening and about 2 hours since I restarted the httpd services, I now only have 59 connections. However, I'm fairly certain based upon the traffic on these 2 websites, that in the evening, there aren't many connections after 2000 hours.
View 3 Replies
View Related
Jan 11, 2010
I need to logrotate logs in directories in /var/log/httpd/.
There are 4 directories in /var/log/httpd/... these directories are /var/log/httpd/access/ /var/log/httpd/debug/ /var/log/httpd/error/ /var/log/httpd/required/
Each of the access, required, error and debug directories have around 20 to 30 access log files of different locations for example:mumbai-access.log, pune-access.log etc..same is the case for 'error' dir 'required' dir and 'debug' dir in /var/log/httpd/
I need to clean up the logfiles in all the 4 directories access, error, debug and required...
I have made a custom logrotate file as follows:
Is the above config correct?
Am I missing something? Will this logrotate the files in /var/log/httpd/access, /var/log/httpd/error, /var/log/httpd/required and /var/log/httpd/error ?
Do i need to include following line in postrotate " /bin/kill -HUP `cat /var/run/httpd.pid 2>/dev/null` 2> /dev/null || true" ?
View 3 Replies
View Related
Jan 27, 2011
I need to change the some configuration in httpd.conf file with out affect any current status of http service.
View 3 Replies
View Related
Jan 18, 2011
Is the logrotate.conf settings global/apply to what is in logrotate.d/? I have olddir /var/log/old_logs in logrotate.conf but logrotate is not placing old rsyslogs in /var/log/old_logs for logrotate.d/rsyslog
View 5 Replies
View Related
Mar 1, 2011
Usually we require vnc to take remote sessions. There was one another i think it was called xdrp or xrdp. I am asking this out of curiosity, is there any way to take remote sessions using http. Like in web conferencing, we invite users to join the conference and then we are able to share desktop. Is there any way to do this on one-to-one basis ? is such a technology exists for linux (for any disto) ?
View 1 Replies
View Related
Mar 22, 2010
Alright, so I upgraded my old fedora 10 server to Fedora 11 with a netinstall CD, but now service httpd restart is broken. I already had to delete the old config file and reinstall apache, but now I can't restart it. I can kill it and then start it manually, but it never stops the running instance and fails to bind the port. I know it's because /var/run/httpd/htpd.pid is in the wrong place, which would be /var/run/httpd.pid, but do I have to make a symlink every time I want to restart it? I edited all the config files to point to the right place, but the system does not honor them. What do I do?
View 3 Replies
View Related
Jul 30, 2010
We are using this url on our testing box and this ip is reverse proxy IP ( 10.192.64.52)
Without reverse proxy:- URL is working fine.
http://10.192.16.77/akc-qa/arsys/ser...TS000000001430
But with reverse proxy its putting escape [%25] in between %3a
[url]
Is it possible to write a rewrite rule to bypass escape [NE] for this particular URL?
View 1 Replies
View Related
Aug 4, 2010
I installed Nagios on my Ubuntu 10.04 server using apt-get and when I accessed the web console, everything was OK. I made some changes to apache (creating some new virtual sites) and since then Nagios gives me a warning message for HTTP with the message, HTTP WARNING: HTTP/1.1 404 Not Found. The sites that I created are working perfectly. I noticed that the attemps are 4/4. Does this need to be reset or does Nagios automatically reset that once it detects the issue is resolved?
View 1 Replies
View Related
Nov 26, 2010
Say that a certain server process generates log files and names them according to the current date; e.g.
server.nov-20.2010.log
server.nov-21.2010.log
server.nov-22.2010.log
server.nov-23.2010.log
i'd like to have logrotate compress the logs that are older than 3 days. Is this possible with logrotate, or do i just schedule a cron job to bzip everything under the folder older than 3 days?
View 1 Replies
View Related
Aug 23, 2010
I have a postfix mailserver that works fine except for the logrotate.
syslog.conf
mail.* -/var/log/mail.log
logrotate.conf
/var/log/mail.log {
[Code]....
So when cron does the logrotate, there is a new logfile but its empty. After i restart the syslogd it gets back to its normal logging.
What am i missing? All this works with CentOS, why is Ubuntu such a pain...
View 3 Replies
View Related
Jun 10, 2011
Running CentOS 5 x64 And today my httpd is running very slow and I can't find a fix. Looked all over different forums
When starting httpd I get the message: /var/lock/subsys/httpd': No space left on device I checked that directory above and there is no file called httpd tried rebooting server
Can't do updates too:
[root@u15438957 ~]# yum update
Loaded plugins: fastestmirror, priorities
rpmdb: unable to join the environment
[Code]...
View 4 Replies
View Related
Mar 18, 2010
I have an Ubuntu server, and I have a special script on logrotate.d to rotate the samba_audit logs:
/var/log/auditsamba/auditoria.log {
weekly
rotate 12
missingok
[code]....
View 14 Replies
View Related
Aug 21, 2010
I've two internet based server ( xx.xx.xx.xx and yy.yy.yy.yy ) The Y server is running VNC server and is responsible for answering to VNC sessions. But I need to hide the IP of Y server so I want X server to be as VNC Proxy and redirect all VNC sessions to Y server.
I guess the best way is to use iptables but actually I can't get it working so
View 4 Replies
View Related
Sep 19, 2010
I have installed fedora 13 in my system. httpd server is also installed. when I tried to start the service of httpd, following error message displayed: Starting httpd: (98)Address already in use: make_sock: could not bind to address [::]:80
View 1 Replies
View Related
Dec 24, 2010
I'm new to Apache configuration so could be misunderstanding but shouldn't the characters highlighted in red below be removed?
Code:
View 9 Replies
View Related
Jun 1, 2009
I have just configured CentOS server my new office. This is the first time I am using CentOS.
Whenever I am trying to change the webroot of the httpd server, it usually say "directory doesn't exist". whereas which can't be the case as it is home directory of the user.
To achieve this, what Do I need to do?
The second issue, I am facing that whenever I am adding new directories to public_html folder, it is not accessible. as the error page 404 is shown. I have changed the permissions to 0705 and ownership is also transferred to other user. What else is required to make new directories accessible by the web?
View 2 Replies
View Related
Mar 23, 2011
I get the following error while starting apache httpd
Code: httpd: Could not determine the server's fully qualified domain name, using <Server's-ip-address> for ServerName I googled it and have come across the following solutions, all of them involve changing the ServerName setting in httpd.conf:ServerName localhost ServerName www.example.com:80 ServerName <ip-address-of-server> ServerName <hostname-of-machine/FQDN>
I am setting up httpd to be accessed from over a LAN, so i don't have a .com domain name. I am thinking of going with the first option, it seems to be working...
View 5 Replies
View Related
Apr 11, 2010
I am trying to do a https access with virtual hosting with apache.
But when I install mod_ssl rpm httpd server does not start again
I am stumped
httpd -t shows syntax is OK
View 5 Replies
View Related
Feb 10, 2010
I have a requirement of using a wildcard certificate for 5 subdomains running under apache httpd server and 1 subdomain under tomcat.Is there any possibility of using the single wildcard certificate both in tomcat and apache
View 3 Replies
View Related
Jan 27, 2011
I've used NoMachine's NX software for some time, but on one of my machines it simply will not work. At first it did, but now when I log in it authenticates, starts downloading a session, but then drops dead. There is no hint of what the problem may be except in /var/log/messages on the server machine:
User 'bill' from '192.168.1.1' logged out. 'NXLogin::reset'
Notice that it's a login reset, after authentication went fine. Researching this I find the problem is often caused by a missing fixed font, which is remedied with an: # apt-get install xfonts-base --reinstall but that doesn't help. I have completely deinstalled and purged the three Nomachine packages on the server, deleting all associated directories, and reinstalled them, but no help. From my client machine I can log into a NX session just fine, so I'm sure it's not the client machine. The Linux version of this software has a limit of 2 sessions, and I suspect the server may have some bogus sessions open, but where? I've done a purge and deleted /usr/NX and /home/bill/.nx so where else could this be?
View 14 Replies
View Related
Nov 9, 2009
accidentally I do something wrong with my server and the httpd folder missing and I need it to setup my mail server and anyone can help me what can I do without reinstalling my Cent OS? Here is the error msg :
[root@mydomain etc]# service httpd stop
Stopping httpd: [FAILED]
[root@mydomain etc]# service httpd restart
[code]....
View 3 Replies
View Related
Jun 14, 2009
I would like to setup 2 linux machine each with two ethernet cards, and need to connect to two subnet e.g. 192.168.*.* and 10.0.*.* . If one linux machine out of order, the other will take over all the services(1 httpd and two named(192.168.*.1 and 10.0.0.1 as server ip). What network config is best and easy to failover to other machine?
View 1 Replies
View Related
Sep 11, 2009
My server keeps freezing up requiring a hard reboot.
CentOS release 5.3 (Final)
httpd-2.2.3-22.el5.centos.2
Here is the error in /var/log/message
Sep 11 00:16:20 localhost kernel: httpd invoked oom-killer: gfp_mask=0x201d2, order=0, oomkilladj=0
Sep 11 00:19:14 localhost kernel: [<c0459e7d>] out_of_memory+0x72/0x1a5
Sep 11 00:19:14 localhost kernel: [<c045b352>] __alloc_pages+0x216/0x297
Sep 11 00:19:14 localhost kernel: [<c045c5bf>] __do_page_cache_readahead+0xc4/0x1c6
Sep 11 00:19:14 localhost kernel: [<c0436d9a>] ktime_get_ts+0x16/0x44
[Code]...
View 5 Replies
View Related
Jul 21, 2011
after installing Debian 6 on a server. When I try to install a software called ActiveCampaign, I get the following error ... "Your server does not appear to be handling sessions properly."I have install Apache, PHP, MySql and Perl already. Also, after the server restarts, I Webmin will not automatically start, even thou it is setup in the Webmin configuration to start with the server. I have to use /etc/init.d/webmin start from a command line after I su.
My last question is about ftp permissions. I have install proftpd and it seems to be working fine, but when I try to edit any file or upload, I can not. In order to upload and manipulate files, I am using WinSCP under root, wich is a big NO NO.Sorry for the three questions, but I figured I would ask all in one post, instead of creating multiple, since I already have your attention.
View 1 Replies
View Related
Oct 5, 2010
Here's the setup: One x86 server (Red Hat 4) with two serial connections to an embedded linux device. One serial connection is to a power control so the device can the restarted, and the other is used for a console to this device. Both serial connections use minicom. People on my team VNC to the x86 server, where they find the two minicom sessions waiting for them as they are left open. I would like to automate connecting to the server and communicating over the serial lines.
Supposing I use ssh to the x86 server, could minicom be used over the serial lines without terminating the minicom processes seen through VNC? If not, is there a way to re-open those windows from within the ssh session, so a user who VNCs to the server won't have to reopen them? Could a single ssh connection control two minicom terminals simultaneously, or would there need to be two ssh connections? It is assumed a user and the automated process will *not* try to access the device at the same time; the automation would trigger during off hours.
View 2 Replies
View Related
Jul 6, 2010
1-what is the shell command which allows me to know how many opened sessions on a mysql server?
2- what are the important points we can manage in a mysql server?
View 2 Replies
View Related
Aug 17, 2011
how do i able to allow some users that are able to create content in directory of http server. For example: i have configured a web server which have default document root /var/www/html, now i want to extend my web server through virtual hosting , i have enable virtual hosting, but i want that user sumit is able to create content in /var/www/html/secret. which is the document root for my virtual site?
View 5 Replies
View Related
Apr 6, 2010
I have been beating my head for the last few weeks on this problem, (although I have been taking the wrong approach, it seems).
I need a gateway to direct web traffic to three separate servers/domains. I have been trying to do this with both a dns server and , (seperatly), apache server to forward requests. The dns server was a no go, and <i can only get apache to redirect http and ftp.
After Googling this ALOT, I believe that what I need is a gateway server to redirect my traffic to the 3 different servers. I have been reading about using using nat and iptables for this and was wondering if anyone had any advice/suggestions on this. The other thought I had was to use something like pfSense to create the gateway, but I am still reading the documentation, and I am unsure if this approach will work.
View 1 Replies
View Related
Nov 24, 2009
I'm running VNC Server on CentOS5 on a brand new installation, this is for some of my students to use specific software via VNC on their machines from home that they otherwise don't have access to on their Windows machines. This has been working fine on our previous server for a few years but the hardware is now outdated so I have upgraded and put a fresh install of CentOS 5 on. This works fine for about 24 hours and then eventually one of the session ends up freezing where the user then can see everything on their session, the cursor can move but they cant click a single thing.
View 1 Replies
View Related