Re: [gentoo-user] raid messages at boot time
Christoph Gysin wrote: >Richard Fish wrote: > > >>I think a far better option would be to filter them in >>/etc/syslog-ng/syslog-ng.conf. Then you do not have to re-patch your >>kernel with every upgrade. >> >> > >This affects only output via syslog. During the md autorun, the kernel hasn't >even finished booting. The post was about output from dmesg, the kernel >ringbuffer. AFAIK there's no way around a kernel patch. > > Ah, yes, you are correct. Personally, I boot my kernel with "quiet", so I don't see any of that 'noise' on startup. I only really care if I am debugging something, so I leave my 'safe' kernel in verbose mode. I suppose you could alias dmesg='dmesg | grep -v "blah blah" | grep -v ...', but that seems a bit insane also. -Richard -- gentoo-user@gentoo.org mailing list
Re: [gentoo-user] raid messages at boot time
Richard Fish wrote: > I think a far better option would be to filter them in > /etc/syslog-ng/syslog-ng.conf. Then you do not have to re-patch your > kernel with every upgrade. This affects only output via syslog. During the md autorun, the kernel hasn't even finished booting. The post was about output from dmesg, the kernel ringbuffer. AFAIK there's no way around a kernel patch. > I use the following instructions to filter out the "PHY reset until link > up" messages that I get when eth0 is "up" (required for ifplugd to > function), but not connected to anything. It can easily be adapted for > the RAID messages. > > # The eth reset messages also bug me... > filter not_eth_reset { > not(match("PHY reset until link up")); > }; > > log { source(src); filter(not_eth_reset); destination(messages); }; Nice example though. Haven't used syslog filtering by now. Christoph -- echo mailto: NOSPAM !#$.'<*>'|sed 's. ..'|tr "<*> !#:2" [EMAIL PROTECTED] -- gentoo-user@gentoo.org mailing list
Re: [gentoo-user] raid messages at boot time
Christoph Gysin wrote: >Patrick wrote: > > >>This is my first raid, i got it working without problems (i think) but my >>dmesg contains this: >>Is this a normal behaviour >> >>md: Autodetecting RAID arrays. >>md: autorun ... >> >> >... > >The md kernel module is quite verbose. Here is a patch to make the kernel print >only the necessary lines: > > I think a far better option would be to filter them in /etc/syslog-ng/syslog-ng.conf. Then you do not have to re-patch your kernel with every upgrade. I use the following instructions to filter out the "PHY reset until link up" messages that I get when eth0 is "up" (required for ifplugd to function), but not connected to anything. It can easily be adapted for the RAID messages. # The eth reset messages also bug me... filter not_eth_reset { not(match("PHY reset until link up")); }; log { source(src); filter(not_eth_reset); destination(messages); }; -Richard -- gentoo-user@gentoo.org mailing list
Re: [gentoo-user] raid messages at boot time
Patrick wrote: > This is my first raid, i got it working without problems (i think) but my > dmesg contains this: > Is this a normal behaviour > > md: Autodetecting RAID arrays. > md: autorun ... ... The md kernel module is quite verbose. Here is a patch to make the kernel print only the necessary lines: --- drivers/md/md.c 2005-03-02 08:37:52.0 +0100 +++ drivers/md/md.c.quiet 2005-04-20 19:53:24.0 +0200 @@ -538,8 +538,10 @@ __u64 ev1, ev2; mdp_super_t *refsb = (mdp_super_t*)page_address(refdev->sb_page); if (!uuid_equal(refsb, sb)) { +/* printk(KERN_WARNING "md: %s has different UUID to %s\n", b, bdevname(refdev->bdev,b2)); +*/ goto abort; } if (!sb_equal(refsb, sb)) { @@ -1035,7 +1037,9 @@ list_add(&rdev->same_set, &mddev->disks); rdev->mddev = mddev; +/* printk(KERN_INFO "md: bind<%s>\n", bdevname(rdev->bdev,b)); +*/ return 0; } @@ -1767,8 +1771,10 @@ MD_BUG(); return; } - + +/* printk(KERN_INFO "md: running: "); +*/ ITERATE_RDEV(mddev,rdev,tmp) { char b[BDEVNAME_SIZE]; @@ -1803,18 +1809,24 @@ mddev_t *mddev; char b[BDEVNAME_SIZE]; +/* printk(KERN_INFO "md: autorun ...\n"); +*/ while (!list_empty(&pending_raid_disks)) { dev_t dev; rdev0 = list_entry(pending_raid_disks.next, mdk_rdev_t, same_set); +/* printk(KERN_INFO "md: considering %s ...\n", bdevname(rdev0->bdev,b)); +*/ INIT_LIST_HEAD(&candidates); ITERATE_RDEV_PENDING(rdev,tmp) if (super_90_load(rdev, rdev0, 0) >= 0) { +/* printk(KERN_INFO "md: adding %s ...\n", +*/ bdevname(rdev->bdev,b)); list_move(&rdev->same_set, &candidates); } @@ -1851,7 +1863,9 @@ mdname(mddev), bdevname(rdev0->bdev,b)); mddev_unlock(mddev); } else { +/* printk(KERN_INFO "md: created %s\n", mdname(mddev)); +*/ ITERATE_RDEV_GENERIC(candidates,rdev,tmp) { list_del_init(&rdev->same_set); if (bind_rdev_to_array(rdev, mddev)) -- echo mailto: NOSPAM !#$.'<*>'|sed 's. ..'|tr "<*> !#:2" [EMAIL PROTECTED] -- gentoo-user@gentoo.org mailing list
[gentoo-user] raid messages at boot time
Hi, This is my first raid, i got it working without problems (i think) but my dmesg contains this: Is this a normal behaviour md: Autodetecting RAID arrays. md: autorun ... md: considering hdb13 ... md: adding hdb13 ... md: hdb12 has different UUID to hdb13 md: hdb11 has different UUID to hdb13 md: hdb10 has different UUID to hdb13 md: hdb9 has different UUID to hdb13 md: hdb8 has different UUID to hdb13 md: hdb7 has different UUID to hdb13 md: hdb6 has different UUID to hdb13 md: hdb5 has different UUID to hdb13 md: hdb1 has different UUID to hdb13 md: adding hda13 ... md: hda12 has different UUID to hdb13 md: hda11 has different UUID to hdb13 md: hda10 has different UUID to hdb13 md: hda9 has different UUID to hdb13 md: hda8 has different UUID to hdb13 md: hda7 has different UUID to hdb13 md: hda6 has different UUID to hdb13 md: hda5 has different UUID to hdb13 md: hda1 has different UUID to hdb13 md: created md9 md: bind md: bind md: running: raid1: raid set md9 active with 2 out of 2 mirrors md: considering hdb12 ... md: adding hdb12 ... md: hdb11 has different UUID to hdb12 ... and so on until md: ... autorun DONE my /proc/mdstat Personalities : [raid1] md1 : active raid1 hdb5[1] hda5[0] 1003904 blocks [2/2] [UU] md2 : active raid1 hdb6[1] hda6[0] 1003904 blocks [2/2] [UU] md3 : active raid1 hdb7[1] hda7[0] 3004032 blocks [2/2] [UU] md4 : active raid1 hdb8[1] hda8[0] 3004032 blocks [2/2] [UU] md5 : active raid1 hdb9[1] hda9[0] 505920 blocks [2/2] [UU] md6 : active raid1 hdb10[1] hda10[0] 505920 blocks [2/2] [UU] md7 : active raid1 hdb11[1] hda11[0] 1003904 blocks [2/2] [UU] md8 : active raid1 hdb12[1] hda12[0] 1003904 blocks [2/2] [UU] md9 : active raid1 hdb13[1] hda13[0] 106125248 blocks [2/2] [UU] md0 : active raid1 hdb1[1] hda1[0] 56128 blocks [2/2] [UU] unused devices: TIA Patrick -- gentoo-user@gentoo.org mailing list
Re: [gentoo-user] raid messages at boot time
Patrick wrote: >Hi, > >This is my first raid, i got it working without problems (i think) but my >dmesg contains this: >Is this a normal behaviour > >md: Autodetecting RAID arrays. >md: autorun ... >md: considering hdb13 ... >md: adding hdb13 ... >md: hdb12 has different UUID to hdb13 > > Looks ok to me... >md: created md9 >md: bind >md: bind >md: running: >raid1: raid set md9 active with 2 out of 2 mirrors > > These are the only lines that really matter. BTW, that is an awful lot of partitions. There is nothing technically wrong with that, but you might find it easier to setup only two RAID1 volumes...one for your root filesystem, and the other with LVM2 for everything else. The advantage of using LVM2 is that you create 'volumes' on the RAID array, rather than a partition on each disk, so you don't have to worry about trying to get the size and type right for each partition. You also get the ability to resize them later if need arises and your filesystem-of-choice supports it. -Richard -- gentoo-user@gentoo.org mailing list
[gentoo-user] raid messages at boot time
Hi, This is my first raid, i got it working without problems (i think) but my dmesg contains this: Is this a normal behaviour md: Autodetecting RAID arrays. md: autorun ... md: considering hdb13 ... md: adding hdb13 ... md: hdb12 has different UUID to hdb13 md: hdb11 has different UUID to hdb13 md: hdb10 has different UUID to hdb13 md: hdb9 has different UUID to hdb13 md: hdb8 has different UUID to hdb13 md: hdb7 has different UUID to hdb13 md: hdb6 has different UUID to hdb13 md: hdb5 has different UUID to hdb13 md: hdb1 has different UUID to hdb13 md: adding hda13 ... md: hda12 has different UUID to hdb13 md: hda11 has different UUID to hdb13 md: hda10 has different UUID to hdb13 md: hda9 has different UUID to hdb13 md: hda8 has different UUID to hdb13 md: hda7 has different UUID to hdb13 md: hda6 has different UUID to hdb13 md: hda5 has different UUID to hdb13 md: hda1 has different UUID to hdb13 md: created md9 md: bind md: bind md: running: raid1: raid set md9 active with 2 out of 2 mirrors md: considering hdb12 ... md: adding hdb12 ... md: hdb11 has different UUID to hdb12 ... and so on until md: ... autorun DONE my /proc/mdstat Personalities : [raid1] md1 : active raid1 hdb5[1] hda5[0] 1003904 blocks [2/2] [UU] md2 : active raid1 hdb6[1] hda6[0] 1003904 blocks [2/2] [UU] md3 : active raid1 hdb7[1] hda7[0] 3004032 blocks [2/2] [UU] md4 : active raid1 hdb8[1] hda8[0] 3004032 blocks [2/2] [UU] md5 : active raid1 hdb9[1] hda9[0] 505920 blocks [2/2] [UU] md6 : active raid1 hdb10[1] hda10[0] 505920 blocks [2/2] [UU] md7 : active raid1 hdb11[1] hda11[0] 1003904 blocks [2/2] [UU] md8 : active raid1 hdb12[1] hda12[0] 1003904 blocks [2/2] [UU] md9 : active raid1 hdb13[1] hda13[0] 106125248 blocks [2/2] [UU] md0 : active raid1 hdb1[1] hda1[0] 56128 blocks [2/2] [UU] unused devices: TIA Patrick -- gentoo-user@gentoo.org mailing list