Product: kav4mailservers-linux-5.5-33
OS: SUSE Linux Enterprise Server 10 SP2 (i586), Releasenotes say it´s not supportet, but the problem is new
Kernel: 2.6.16.60-0.42.5-bigsmp #1 SMP Mon Aug 24 09:41:41 UTC 2009 i686 i686 i386 GNU/Linux
Hi!
We do not really know how long this problems exists but when we make a trace we see that the process needs more than 10 minutes to analyse the rollback.conf file.
11:02:10 open("/var/db/kav/5.5/kav4mailservers/bases.backup/rollback.conf", O_RDONLY|O_NONBLOCK) = 4
11:02:10 lseek(4, 0, SEEK_END) = 571266
11:02:10 lseek(4, 0, SEEK_SET) = 0
11:02:10 mmap2(NULL, 573440, PROT_READ|PROT_WRITE, MAP_PRIVATE|MAP_ANONYMOUS, -1, 0) = 0xb7d7c000
11:02:10 read(4, "\n\n[RecentChangedFiles]\nentry0=0("..., 571266) = 571266
11:02:10 close(4) = 0
11:13:42 write(1, "Checking product license keys\n", 30) = 30
11:13:42 write(1, "Initializing\n", 13) = 13
11:13:42 time(NULL) = 1257329622
11:13:42 access("/var/db/kav/5.5/kav4mailservers/bases/updcfg.xml", R_OK) = 0
11:13:42 open("/var/db/kav/5.5/kav4mailservers/bases/updcfg.xml", O_RDONLY|O_LARGEFILE) = 4
11:13:42 read(4, "<?xml version=\"1.0\"?>\r\n\r\n<Update"..., 8192) = 8192
So much time and so many CPU cycles just for the rollback.conf? I can´t believe that this is OK.
After deleting all bases-files the first update runs fast the following need the 10 minutes before contacting the update server.
We have more than one systems and on all we have the problem. They are no VMWares, they have enough RAM, CPU-Power and while the keepup2date runs with 100% IO-Wait is nearly 0%.
Any Ideas? Any help would be appreaciated. Ff course ;-)
Matthias