TheKorn2
March 26th, 2009, 10:21 AM
I'm new to RAID, so apologies in advance if this question is all naff, but here goes. I have four 1T hard drives. I'm using the alternate setup disc for 9.0 alpha 6, i386.
I set up one 10G partition on each drive, and raided them together in a raid-1 array. That's my boot (and root) partition. 2 active, 2 standby (or whatever the installer calls them). That seems to have gone OK. (ext3fs, in case anyone cares.)
I also set up a 1G swap partition on each drive. These are NOT raided together, they're just single swap partitions on each drive that the kernel can figure out how it wants to use them. :)
Finally, the remainder of the space on each drive I set up as four-way raid 5 partitions, and I'm using jfs for the file system.
That *seems* to have gone all right. Jaunty Jackhole installed OK (despite the video corruption in the text mode installer), and I later added an entry to fstab to mount the 3TB raid-5 partition by UUID.
~$ mount
/dev/md0 on / type ext3 (rw,relatime,errors=remount-ro)
(snip)
/dev/md1 on /mythtv type jfs (rw)
(snip)
What I don't understand is that I now have two processes that come up whenever I start the system, and are consuming 30% of the CPU even when it's idle:
top - 09:12:52 up 2:50, 3 users, load average: 1.31, 1.38, 1.51
Tasks: 134 total, 3 running, 131 sleeping, 0 stopped, 0 zombie
Cpu(s): 0.7%us, 22.3%sy, 1.5%ni, 68.1%id, 2.7%wa, 0.7%hi, 3.8%si, 0.0%st
Mem: 3224096k total, 782064k used, 2442032k free, 69032k buffers
Swap: 3919824k total, 0k used, 3919824k free, 415992k cached
PID USER PR NI VIRT RES SHR S %CPU %MEM TIME+ COMMAND
1504 root 15 -5 0 0 0 R 26 0.0 56:42.96 md1_raid5
1506 root 15 -5 0 0 0 D 6 0.0 15:13.70 md1_resync
Is this normal for a raid5 array to immediately go into resync on boot? I don't think so, but it IS a new array and I'm not well versed in how mdadm handles raid arrays. (Maybe it's setting something up? Maybe?) It's been doing this for a few hours now!
The hard drive controller light is continuously on. The system is responsive, but throwing 30% of the CPU away right off the top while doing nothing *bugs* me.
But here's the weird thing... if I run iotop, nothing is going out to the hard drives!
Total DISK READ: 0 B/s | Total DISK WRITE: 0 B/s
PID USER DISK READ DISK WRITE SWAPIN IO> COMMAND
1 root 0 B/s 0 B/s 0.00 % 0.00 % init
2 root 0 B/s 0 B/s 0.00 % 0.00 % [kthreadd]
So what's going on here? Is this normal? A bug in jaunty? Should I be concerned? Will the system eventually figure out my raid 5 is OK and knock it off? Do I just have to 'wait it out'?
Thanks!!
(BTW, I did some very preliminary reading/writing to the jfs partition and that seemed to go OK. So it's responsive, but something weird is going on here, I think.)
I set up one 10G partition on each drive, and raided them together in a raid-1 array. That's my boot (and root) partition. 2 active, 2 standby (or whatever the installer calls them). That seems to have gone OK. (ext3fs, in case anyone cares.)
I also set up a 1G swap partition on each drive. These are NOT raided together, they're just single swap partitions on each drive that the kernel can figure out how it wants to use them. :)
Finally, the remainder of the space on each drive I set up as four-way raid 5 partitions, and I'm using jfs for the file system.
That *seems* to have gone all right. Jaunty Jackhole installed OK (despite the video corruption in the text mode installer), and I later added an entry to fstab to mount the 3TB raid-5 partition by UUID.
~$ mount
/dev/md0 on / type ext3 (rw,relatime,errors=remount-ro)
(snip)
/dev/md1 on /mythtv type jfs (rw)
(snip)
What I don't understand is that I now have two processes that come up whenever I start the system, and are consuming 30% of the CPU even when it's idle:
top - 09:12:52 up 2:50, 3 users, load average: 1.31, 1.38, 1.51
Tasks: 134 total, 3 running, 131 sleeping, 0 stopped, 0 zombie
Cpu(s): 0.7%us, 22.3%sy, 1.5%ni, 68.1%id, 2.7%wa, 0.7%hi, 3.8%si, 0.0%st
Mem: 3224096k total, 782064k used, 2442032k free, 69032k buffers
Swap: 3919824k total, 0k used, 3919824k free, 415992k cached
PID USER PR NI VIRT RES SHR S %CPU %MEM TIME+ COMMAND
1504 root 15 -5 0 0 0 R 26 0.0 56:42.96 md1_raid5
1506 root 15 -5 0 0 0 D 6 0.0 15:13.70 md1_resync
Is this normal for a raid5 array to immediately go into resync on boot? I don't think so, but it IS a new array and I'm not well versed in how mdadm handles raid arrays. (Maybe it's setting something up? Maybe?) It's been doing this for a few hours now!
The hard drive controller light is continuously on. The system is responsive, but throwing 30% of the CPU away right off the top while doing nothing *bugs* me.
But here's the weird thing... if I run iotop, nothing is going out to the hard drives!
Total DISK READ: 0 B/s | Total DISK WRITE: 0 B/s
PID USER DISK READ DISK WRITE SWAPIN IO> COMMAND
1 root 0 B/s 0 B/s 0.00 % 0.00 % init
2 root 0 B/s 0 B/s 0.00 % 0.00 % [kthreadd]
So what's going on here? Is this normal? A bug in jaunty? Should I be concerned? Will the system eventually figure out my raid 5 is OK and knock it off? Do I just have to 'wait it out'?
Thanks!!
(BTW, I did some very preliminary reading/writing to the jfs partition and that seemed to go OK. So it's responsive, but something weird is going on here, I think.)