Linux Blog

Getting the following bug…. bad luck :(
Running Centos, 32 bit

BUG: warning at drivers/net/e1000e/ich8lan.c:361/e1000_acquire_swflag_ich8lan() (Not tainted)
[] e1000_acquire_swflag_ich8lan+0×6e/0×86 [e1000e]
[] e1000e_read_kmrn_reg+0×10/0×4e [e1000e]
[] e1000e_gig_downshift_workaround_ich8lan+0×23/0×66 [e1000e]
[] e1000_intr_msi+0×46/0xd4 [e1000e]
[] handle_IRQ_event+0×23/0×49
[] __do_IRQ+0×84/0xd6
[] do_IRQ+0×93/0xae
[] common_interrupt+0×1a/0×20
[] e1000e_mc_addr_list_update_generic+0×95/0×148 [e1000e]
[] e1000_mc_addr_list_update+0xe/0×12 [e1000e]
[] e1000_set_multi+0xcd/0xd5 [e1000e]
[] __dev_mc_upload+0×1d/0×1e
[] dev_mc_upload+0×28/0×3b
[] dev_open+0×44/0×62
[] dev_change_flags+0×47/0xe4
[] devinet_ioctl+0×24e/0×542
[] sock_ioctl+0×191/0×1b3
[] sock_ioctl+0×0/0×1b3
[] do_ioctl+0×1c/0×5d
[] vfs_ioctl+0×24a/0×25c
[] sys_ioctl+0×48/0×5f
[] syscall_call+0×7/0xb
=======================
pci0000:00: eth0: Link is Up 100 Mbps Full Duplex, Flow Control: RX
pci0000:00: eth0: 10/100 speed: disabling TSO

Many of our readers are asking what KeepAliveTimeout setting should be used for Apache config. We usually select 2 to 5 seconds as it provides best performance for sites with medium visitors and prefer to use Nginx or any other front-end proxy to better manage thousands of multiple concurrent users.

We have seen great performance on using ReiserFS for mail system spools and in our tests ReiserFS file system spool for Postfix queues was very fast. For best performance we used

notail, noatime mount options in /etc/fstab file.

Full example:

/dev/sda2/usr reiserfs notail,noatime,rw,auto 0 0

notatime – turns off atimes calls, reduces IO and greatly increases file system performance
notail – increases performance of ReiserFS especially for small files – works great for mail servers, web servers and caches. This option disables file pckaging in tree.