Hi Alexandre,

Alexandre Correa wrote:
How many users is your Squid box serving? It's strange that there is no
errors. Where have you defined the cache_log directive in your squid.conf?

+- 300~400 users simultaneously

cache_access_log none
cache_log /var/log/squid/cache.log
cache_store_log none

cache_dir aufs /var/spool/squid 110000 72 256
cache_mem 256 MB

That's a 110 GB cache which is big. So that is absolutely nothing there in /var/log/squid/cache.log before your Squid process goes Zombie?



When you say "after some time running .. squid proccess refusing
connections", is it a few minutes or hours or even days?
an accurate time does not exist, after some hours, when the load increases ..
Can you post the output of "squidclient mgr:info"?

Posting your squid.conf may help too.

maresia# squidclient mgr:info
Squid Object Cache: Version 2.6.STABLE16
Start Time:     Fri, 02 Nov 2007 04:43:44 GMT
Current Time:   Fri, 02 Nov 2007 07:24:54 GMT
Connection information for squid:
        Number of clients accessing cache:      0
        Number of HTTP requests received:       28162
        Number of ICP messages received:        21
        Number of ICP messages sent:    21

So you also have sibling caches too? But why is there so little ICP traffic compared to HTTP traffic?


        Number of queued ICP replies:   0
        Request failure ratio:   0.00
        Average HTTP requests per minute since start:   100.2
        Average ICP messages per minute since start:    0.1
        Select loop called: 795855 times, 21.196 ms avg
Cache information for squid:
        Request Hit Ratios:     5min: 48.1%, 60min: 43.4%
        Byte Hit Ratios:        5min: 4.5%, 60min: 6.6%
        Request Memory Hit Ratios:      5min: 2.2%, 60min: 1.2%
        Request Disk Hit Ratios:        5min: 50.0%, 60min: 26.6%
        Storage Swap size:      2980474 KB
        Storage Mem size:       50900 KB
        Mean Object Size:       17.57 KB
        Requests given to unlinkd:      0
Median Service Times (seconds)  5 min    60 min:
        HTTP Requests (All):   0.08265  0.09736
        Cache Misses:          0.39928  0.37825
        Cache Hits:            0.00286  0.00463
        Near Hits:             0.09219  0.09219
        Not-Modified Replies:  0.00286  0.00000
        DNS Lookups:           0.10428  0.17826
        ICP Queries:           0.00000  0.00000
Resource usage for squid:
        UP Time:        16869.319 seconds
        CPU Time:       85.383 seconds
        CPU Usage:      0.51%
        CPU Usage, 5 minute avg:        0.38%
        CPU Usage, 60 minute avg:       0.28%
        Process Data Segment Size via sbrk(): 154172 KB
        Maximum Resident Size: 159328 KB
        Page faults with physical i/o: 1
Memory accounted for:
        Total accounted:        75685 KB
        memPoolAlloc calls: 4466729
        memPoolFree calls: 4009299
File descriptor usage for squid:
        Maximum number of file descriptors:   4096
        Largest file desc currently in use:     26
        Number of file desc currently in use:   21
        Files queued for open:                   0
        Available number of file descriptors: 4075
        Reserved number of file descriptors:   100
        Store Disk files open:                   0
        IO loop method:                     kqueue
Internal Data Structures:
        169850 StoreEntries
          5569 StoreEntries with MemObjects
          5564 Hot Object Cache Items
        169667 on-disk objects


at this moment i have less traffic/users .. i will run later when load
increase.. (if squid not freeze before).. it´s very strange..

When your Squid process gets freeze or goes zombie, what errors does your web browser gives you? I guess that machine does not loose network connectivity?




kernel compiled with options: (6.2-RELEASE-p8 FreeBSD)


options         SYSVSHM                 # SYSV-style shared memory
options         SYSVMSG                 # SYSV-style message queues
options         SYSVSEM                 # SYSV-style semaphores

options MSGMNB=16384
options MSGMNI=41
options MSGSEG=2049
options MSGSSZ=64
options MSGTQL=512
options SHMSEG=16
options SHMMNI=32
options SHMMAX=2097152
options SHMALL=3096

When your Squid process goes to a zombie state again, check the number of mbufs used at that very moment.

Maybe, just maybe, it could be related to your mbufs running out.



thanks !!

Thanking you...



regards,

AlexandrE




# ps auwx | grep squid

USER        PID %CPU %MEM   VSZ   RSS  TT  STAT STARTED      TIME COMMAND
squid       807  0.0 16.1 679548 671268  ??  T    10:52PM  10:50.02 (squid)
-D -s (squid)
squid       864  0.0  0.0  2472   752  ??  Is   10:52PM   0:00.00 (unlinkd)
(unlinkd)
root      36341  0.0  0.0  5852  1212  p0  R+   10:59AM   0:00.00 grep squi=
d

using AUFS !!

squid 2.6.STABLE16
configure options:

 '--program-prefix=3D' '--prefix=3D/usr' '--exec-prefix=3D/usr'
'--bindir=3D/usr/bin' '--sbindir=3D/usr/sbin' '--sysconfdir=3D/etc'
'--includedir=3D/usr/include' '--libdir=3D/usr/lib' '--libexecdir=3D/usr/li=
bexec'
'--sharedstatedir=3D/usr/com' '--mandir=3D/usr/share/man'
'--infodir=3D/usr/share/info' '--exec_prefix=3D/usr' '--bindir=3D/usr/sbin'
'--libexecdir=3D/usr/lib/squid' '--localstatedir=3D/var'
'--sysconfdir=3D/etc/squid' '--disable-useragent-log' '--disable-referer-lo=
g'
'--enable-kqueue' '--enable-snmp' '--enable-removal-policies=3Dheap,lru'
'--enable-storeio=3Daufs,coss,diskd,ufs' '--enable-ssl'
'--enable-ipf-transparent' '--enable-linux-netfilter' '--with-pthreads'
'--disable-dependency-tracking' '--enable-cachemgr-hostname=3Dlocalhost'
'--disable-ident-lookups' '--enable-underscores' '--datadir=3D/usr/share'
'--with-maxfd=3D4096' '--enable-async-io' '--disable-dlmalloc' '--with-aio'

somebody knowns wat=B4s happens ?!
Is that the output of "squid -v"? By the way, what is the 3D reference
in your compilation parameters for?

Did you install Squid from FreeBSD ports or did you compile it yourself?


thanks..

regards,


--

With best regards and good wishes,

Yours sincerely,

Tek Bahadur Limbu

System Administrator

(TAG/TDG Group)
Jwl Systems Department

Worldlink Communications Pvt. Ltd.

Jawalakhel, Nepal

http://www.wlink.com.np

http://teklimbu.wordpress.com






--

With best regards and good wishes,

Yours sincerely,

Tek Bahadur Limbu

System Administrator

(TAG/TDG Group)
Jwl Systems Department

Worldlink Communications Pvt. Ltd.

Jawalakhel, Nepal

http://www.wlink.com.np

http://teklimbu.wordpress.com

Reply via email to