Skip to main content

repeating nscd monitoring messages after update to 11.52.0.14

Comments

25 comments

  • DJPRMF
    I notice this also in some of my servers. The nscd was installed in the latest update... I don't know if is safe to disable/stop it, but is filling the /var/log/messages almost every second...
    0
  • cPanelMichael
    Hello :) It's by design that NSCD is enabled on systems during the update to cPanel version 11.52. Here's some more information from the 11.52 Change Log - Documentation - cPanel Documentation Thank you.
    0
  • DJPRMF
    Adding a little of info given by the cPanel support, the NSCD service can be disabled in the WHM >> Service Manager >> Name Service Cache Daemon . Should be safe to disable while there is no solution for the problem...
    0
  • ladydi711
    Thank you both for the info!!!
    0
  • Kobor
    I came here to see why suddenly i have nscd logging every 10 sec in my logs, and to my surprise cPanel is responsible again. It seems to me that a lot of changes were made in 11.52 without much thinking, for example Exim 10 sec delay on SMTP connections, now this. I don't really like when i have a working configured system, then over night the behavior of server changes ( like Exim change. or on all servers the log is full of nscd messages ). Especially since both example i used is pretty much useless IMHO. Like i don't really get how exactly will help exim, when exim pretty much doesn't user /etc/{passwd,shadow}, the hosts file contains one or maybe two line on 90% of the cpanel servers, and pretty much nobody uses netgroups ( Nscd is configured to cache these 4 files ). And you may argument that it saves a tiny bit of microsecond, but we loose anyway much more with RBL checks, reverse DNS lookups, virus scanning, spam scanning . So pretty much nscd is useless. I'm handling and handled a couple big servers with considerable email traffic, but passwd/shadow file scanning was never the bottleneck. I had a great respect over the years for the cpanel devs, but the past year you started loosing that. Seems changes are pushed from people who never worked on real live systems.
    0
  • Hedloff
    So nscd is Name Service Cache Daemon and not cPanel DNS Admin Cache. Why have you enabled this automatically for all our servers after update to 11.52? What does it do? I see it's taking some cpu/resources and logging all the time as mentioned earlier. We have disabled cPanel DNS Admin Cache on the servers after testing that. It only caused issues. Now it seems like we have to disable Name Service Cache Daemon also since it's just causing issues and taking resources. I totally agree with you Kobor. cPanel is moving in the wrong direction the last 2-3 years. Only crappy features are released! Nothing that is even usefull. And that you set them as active without letting customers to choose seems for me as pretty fu**ed up!
    0
  • did-vmonroig
    We've same problem. NSCD is flooding /var/log/messages:
    root@server [~]# tail -f /var/log/messages Oct 27 17:35:40 server nscd: 32707 monitoring file `/etc/passwd` (1) Oct 27 17:35:40 server nscd: 32707 monitoring directory `/etc` (2) Oct 27 17:35:40 server nscd: 32707 monitoring file `/etc/group` (3) Oct 27 17:35:40 server nscd: 32707 monitoring directory `/etc` (2) Oct 27 17:35:40 server nscd: 32707 monitoring file `/etc/passwd` (1) Oct 27 17:35:40 server nscd: 32707 monitoring directory `/etc` (2) Oct 27 17:35:40 server nscd: 32707 monitoring file `/etc/group` (3) Oct 27 17:35:40 server nscd: 32707 monitoring directory `/etc` (2) Oct 27 17:35:40 server nscd: 32707 monitoring file `/etc/passwd` (1) Oct 27 17:35:40 server nscd: 32707 monitoring directory `/etc` (2) Oct 27 17:35:41 server nscd: 32707 monitoring file `/etc/group` (3) Oct 27 17:35:41 server nscd: 32707 monitoring directory `/etc` (2) Oct 27 17:35:41 server nscd: 32707 monitoring file `/etc/group` (3) Oct 27 17:35:41 server nscd: 32707 monitoring directory `/etc` (2) Oct 27 17:35:41 server nscd: 32707 monitoring file `/etc/passwd` (1) Oct 27 17:35:41 server nscd: 32707 monitoring directory `/etc` (2) Oct 27 17:35:41 server nscd: 32707 monitoring file `/etc/passwd` (1) Oct 27 17:35:41 server nscd: 32707 monitoring directory `/etc` (2) Oct 27 17:35:42 server nscd: 32707 monitoring file `/etc/group` (3) Oct 27 17:35:42 server nscd: 32707 monitoring directory `/etc` (2) Oct 27 17:35:42 server nscd: 32707 monitoring file `/etc/passwd` (1) Oct 27 17:35:42 server nscd: 32707 monitoring directory `/etc` (2)
    We're going to disable the service while cPanel provides a solution.
    0
  • cPanelMichael
    Now it seems like we have to disable Name Service Cache Daemon also since it's just causing issues and taking resources

    Could you elaborate on the additional issues, beyond the excessive logging in /var/log/messages so we can investigate those problems? Thank you.
    0
  • DJPRMF
    nscd will fill the logs pretty fast.... Even in is own log file (like what cpanel team is trying to do apparently), those entries (probably) will still be created, wasting CPU and I/O.... Don't to mention the log file size, that can be big (just search in the google and will find many reports of the nscd logfile HUGE sizes, or imagine those entries being replicated over and over again in another log). I don't know if adding the nscd will help anything, comparing to the disadvantages...
    0
  • feanorknd
    Hi... I have seen /etc/nscd.conf is being used, as I tested paranoia mode to yes and it is ok at "nscd -g". But if I try to enable "logfile /var/log/nscd.log", this is not running, because all debug messages are being directed to /var/log/messages in spite of any file specified at logfile line. My intention is to redirect logfile to /dev/null directly ;) But there is no way... I have also checked that nscd user has privileges at /var/log/nscd.log file, but no success... There is almost documentation around this online... I cannot understand why nscd is always printing logs at /var/log/messages in spite of configuring logfile. I think /dev/null logfile is a great idea for nscd.
    0
  • feanorknd
    Ok........ fixed..... I tried this: Due to documentation: logfile file Specifies the name of the debug log-file that nscd(8) should use if debug-level is higher than 0. If this option is not set, nscd(8) will write its debug output to stderr. debug-level level If level is higher than 0, nscd(8) will create some debug output. The higher the level, the more verbose the output.
    So at nscd.conf..... ... logfile /dev/null debug-level 1 ...
    Result is nscd is sending log messages to /dev/null.... I think now there is not any IO or extra CPU proccessing... optimal? What do you think? :rolleyes: But...... I keep preffering using own named service to cache DNS queries... At /etc/resolv.conf nameserver 127.0.0.1 nameserver 213.133.99.99
    If named at 127.0.0.1 fails, there is always a second nameserver for queries... Now, at named.conf, allow recursion only for internal and disable for external. Your named service will cache all queries done, and if not hit, recursion to the parent root servers to resolv. In my case, it is running perfectly since years ago.
    0
  • AnthonyFrancis
    I have just noticed this on several hosts that have been recently updated. nscd is a noisy beast. It would be awesome if I didn't have to go into hundreds of hosts and kill it to make logs readable again.
    0
  • popeye
    We get this message most days this is another reason we are moving away from cpanel, because they keep doing pointless updates.
    0
  • Serra
    I have just noticed this on several hosts that have been recently updated. nscd is a noisy beast. It would be awesome if I didn't have to go into hundreds of hosts and kill it to make logs readable again.

    I was able to turn off the noise on my logs, then a couple of days later, it came back. Thanks....
    0
  • cPanelMichael
    cPanel version 11.52.1.0 is now available on the "Current" build tier and includes an improvement to this issue: Fixed case CPANEL-1966: Set nscd to log to /var/log/nscd.log. Thank you.
    0
  • kernow
    Am still getting this problem, running WHM 11.52.1 (build 2) /var/log/nscd.log exists but nothing gets written to it.
    0
  • cPanelMichael
    Am still getting this problem, running WHM 11.52.1 (build 2) /var/log/nscd.log exists but nothing gets written to it.

    Hello :) To clarify, are you stating that information is still logged to /var/log/messages on your system, or is the issue now that no data is logged to /var/log/nscd.log? Did you make any previous changes to the /etc/nscd.conf file? Thank you.
    0
  • kernow
    Correct, nscd information is still logged to /var/log/messages. No manual changes have been made.
    0
  • cPanelMichael
    Correct, nscd information is still logged to /var/log/messages. No manual changes have been made.

    Internal case CPANEL-2581 is open to report this behavior (on systems with the resolution from CPANEL-1966) to our developers. I will update this thread with more information as it becomes available. Thank you.
    0
  • kernow
    Thanks for the update :) We estimate that NSCD is adding around 3 to 4,000 lines an hour to /var/log/messages so we have disabled the NSCD daemon until you have a fix for it.
    0
  • blalow
    indeed! Here is how to turn it off /etc/init.d/nscd stop chkconfig --levels 2345 nscd off Not sure if this service is "monitored". I will see if my phone gets a hundred texts :)
    0
  • kernow
    By default it is monitored, it can be disabled in >>>WHM>>>Service Manager>>>Name Service Cache Daemon
    0
  • feanorknd
    Having this stats, I preffer to disable nscd service, sincerely... In previous answer from me, I did tell how to configure nscd.conf to disable logging to messages by sending logs to /dev/null... but I also have seen nscd frequently at top, using considerable resources, so this is useless at all. nscd -g configuraci"n nscd: 1 nivel de depuraci"n del servidor 4d 13h 7m 0s tiempo de funcionamiento del servidor 17 current number of threads 32 maximum number of threads 13 number of times clients had to wait no paranoia mode enabled 3600 restart internal 5 reload count passwd cache: yes cache is enabled yes cache is persistent yes cache is shared 211 suggested size 216064 total data pool size 1368 used data pool size 600 seconds time to live for positive entries 20 seconds time to live for negative entries 1464371 cache hits on positive entries 437746 cache hits on negative entries 7259857 cache misses on positive entries 59600 cache misses on negative entries 20% cache hit rate 14 current number of cached values 471 maximum number of cached values 10 maximum chain length searched 21 number of delays on rdlock 27 number of delays on wrlock 0 memory allocations failed yes check /etc/passwd for changes group cache: yes cache is enabled yes cache is persistent yes cache is shared 211 suggested size 1264347 total data pool size 1133120 used data pool size 3600 seconds time to live for positive entries 60 seconds time to live for negative entries 7959 cache hits on positive entries 9107 cache hits on negative entries 4267214 cache misses on positive entries 670086 cache misses on negative entries 0% cache hit rate 7 current number of cached values 486 maximum number of cached values 10 maximum chain length searched 10 number of delays on rdlock 12 number of delays on wrlock 0 memory allocations failed yes check /etc/group for changes hosts cache: yes cache is enabled yes cache is persistent yes cache is shared 211 suggested size 216064 total data pool size 31064 used data pool size 7200 seconds time to live for positive entries 60 seconds time to live for negative entries 13380 cache hits on positive entries 1891 cache hits on negative entries 142242 cache misses on positive entries 617074 cache misses on negative entries 1% cache hit rate 234 current number of cached values 479 maximum number of cached values 15 maximum chain length searched 12 number of delays on rdlock 0 number of delays on wrlock 0 memory allocations failed yes check /etc/hosts for changes services cache: yes cache is enabled yes cache is persistent yes cache is shared 211 suggested size 216064 total data pool size 544 used data pool size 28800 seconds time to live for positive entries 20 seconds time to live for negative entries 0 cache hits on positive entries 0 cache hits on negative entries 68 cache misses on positive entries 26 cache misses on negative entries 0% cache hit rate 5 current number of cached values 15 maximum number of cached values 1 maximum chain length searched 0 number of delays on rdlock 0 number of delays on wrlock 0 memory allocations failed yes check /etc/services for changes netgroup cache: no cache is enabled no cache is persistent no cache is shared 0 suggested size 0 total data pool size 0 used data pool size 28800 seconds time to live for positive entries 20 seconds time to live for negative entries 0 cache hits on positive entries 0 cache hits on negative entries 0 cache misses on positive entries 0 cache misses on negative entries 0% cache hit rate 0 current number of cached values 0 maximum number of cached values 0 maximum chain length searched 0 number of delays on rdlock 0 number of delays on wrlock 0 memory allocations failed yes check /etc/netgroup for changes
    0
  • gryzli
    We too have struggled with the "nscd madness". ======== Guys(cPanel ones), enabling those kind of services "by default" while they weren't running is a really "crap deployment behaviour". This service almost killed multiple hypervisors with bunch of cPanel-based VPS's. ======== Given the fact that the service is "monitored" by default and /etc/init.d/nscd stop is ineffective, please provide solution for: - Disabling NSCD and disabling it from tailwatchd, with command line only tools If anybody else needs to "fast disable" nscd on a hundred/thousands of servers, here is what I've done (until cPanel provide "good way of disabling it"):
    mv /usr/sbin/nscd{,.bak} pkill -9 -f nscd
    Update: The right way to disable NSCD (suggested by cPanel support)
    # This disables nscd in Tailwatchd touch /etc/nscddisable ; /scripts/restartsrv nscd # This disables nscd in init.d chkconfig off nscd
    0
  • cPanelMichael
    Hello :) To update, a resolution associated with internal case CPANEL-2581 is included in cPanel version 54: Fixed case CPANEL-2581: Suppressed log noise from NSCD by sending it to /dev/null. cPanel version 54 is currently available in the "Current" and "Edge" build tiers: cPanel & WHM Versions and the Release Process - cPanel Knowledge Base - cPanel Documentation Thank you.
    0

Please sign in to leave a comment.