<div dir="ltr"><div>ok thank you.</div>
<div> </div>
<div>You mentioned the kernel being old, which kernel would you recommend at this point?<br><br></div>
<div class="gmail_quote">On Thu, Sep 24, 2009 at 8:42 PM, Sunil Mushran <span dir="ltr">&lt;<a href="mailto:sunil.mushran@oracle.com">sunil.mushran@oracle.com</a>&gt;</span> wrote:<br>
<blockquote class="gmail_quote" style="PADDING-LEFT: 1ex; MARGIN: 0px 0px 0px 0.8ex; BORDER-LEFT: #ccc 1px solid">Then remove (temporarily) the node from the cluster. You don&#39;t want one<br>node to negatively affect the functioning of the rest.<br>

<br>The reason we recommend forcing a reset on oops is because we cannot<br>predict its effect on the cluster. Because the oops could be in any<br>component in the kernel. Sticking to ocfs2, say if dlm_thread oopses.<br>
Well, then the node would be unable to respond to dlm messages. The<br>
cluster would grind to a halt. If reset was enabled, the other 9 would<br>pause, recover the dead node and continue working. The dead node<br>would reset and then rejoin the cluster.<br><br>In your specific case, it could be harmless. But I wouldn&#39;t bet on it.<br>

<br>Laurence Mayer wrote:<br>
<blockquote class="gmail_quote" style="PADDING-LEFT: 1ex; MARGIN: 0px 0px 0px 0.8ex; BORDER-LEFT: #ccc 1px solid">
<div class="im">ok will do.<br>Just a little background:<br>We are doing reads of up to 220MB/s for 20min (aggregated on all 10 nodes) and towards the end of the 20min we are<br>writing ~45 x 2k files to the OCFS2 volume. During the read, I notice that the Cache Buffers on all the nodes are exhausted .<br>

 This oops only happens currently on one of the nodes. I am relucatnt to force a reboot on oops.<br>Is this a must?<br> Thanks<br>Laurence<br> <br><br></div>
<div class="im"> On Thu, Sep 24, 2009 at 8:06 PM, Sunil Mushran &lt;<a href="mailto:sunil.mushran@oracle.com" target="_blank">sunil.mushran@oracle.com</a> &lt;mailto:<a href="mailto:sunil.mushran@oracle.com" target="_blank">sunil.mushran@oracle.com</a>&gt;&gt; wrote:<br>

<br>   So a read on some file on a xfs volume, triggered a mem alloc<br>   which inturn<br>   triggered the kernel to free up some memory. The oops happens when<br>   it is<br>   trying to free up an ocfs2 inode.<br><br>
   Do:<br>
   # cat /proc/sys/kernel/panic_on_oops<br><br>   If this returns 0, do:<br>   # echo 1 &gt; /proc/sys/kernel/panic_on_oops<br>   This is documented in the user&#39;s guide.<br><br>   File a bugzilla in <a href="http://oss.oracle.com/bugzilla" target="_blank">oss.oracle.com/bugzilla</a><br>

</div>   &lt;<a href="http://oss.oracle.com/bugzilla" target="_blank">http://oss.oracle.com/bugzilla</a>&gt;. _Attach_ this oops report. 
<div>
<div></div>
<div class="h5"><br>   Do not cut-paste. It is hard to read. Also _attach_ the objdump<br>   output.<br>   # objdump -DSl /lib/modules/`uname -r`/kernel/fs/ocfs2/ocfs2.ko<br>   &gt;/tmp/ocfs2.out<br><br>   Bottomline, that it is working just means that you will encounter<br>

   the problem<br>   later. The problem in this case will most likely be another oops.<br>   Or, a hang.<br><br>   Upload the outputs. I&#39;ll try to see if we have already addressed<br>   this issue.<br>   This kernel is fairly old, btw.<br>

<br>   Sunil<br><br>   Laurence Mayer wrote:<br><br>       OS: Ubuntu 8.04 x64<br>       Kern: Linux n1 2.6.24-24-server #1 SMP Tue Jul 7 19:39:36 UTC<br>       2009 x86_64 GNU/Linux<br>       10 Node Cluster<br>       OCFS2 Version:<br>

       ocfs2-tools                      1.3.9-0ubuntu1                                  ocfs2-tools-static-dev    1.3.9-0ubuntu1                            ocfs2console                  1.3.9-0ubuntu1                                      root@n1:~# cat /proc/meminfo<br>

       MemTotal:     16533296 kB<br>       MemFree:         47992 kB<br>       Buffers:        179240 kB<br>       Cached:       13185084 kB<br>       SwapCached:         72 kB<br>       Active:        4079712 kB<br>       Inactive:     12088860 kB<br>

       SwapTotal:    31246416 kB<br>       SwapFree:     31246344 kB<br>       Dirty:            2772 kB<br>       Writeback:           4 kB<br>       AnonPages:     2804460 kB<br>       Mapped:          51556 kB<br>       Slab:           223976 kB<br>

       SReclaimable:    61192 kB<br>       SUnreclaim:     162784 kB<br>       PageTables:      12148 kB<br>       NFS_Unstable:        8 kB<br>       Bounce:              0 kB<br>       CommitLimit:  39513064 kB<br>       Committed_AS:  3698728 kB<br>

       VmallocTotal: 34359738367 kB<br>       VmallocUsed:     53888 kB<br>       VmallocChunk: 34359684419 kB<br>       HugePages_Total:     0<br>       HugePages_Free:      0<br>       HugePages_Rsvd:      0<br>       HugePages_Surp:      0<br>

       Hugepagesize:     2048 kB<br><br>       I have started seeing the below on one of the nodes. The node<br>       does not reboot it continues to function &quot;normally&quot;<br><br>       Is this a memory issue?<br>

<br>       Please can you provide direction.<br><br><br>       Sep 24 16:31:46 n1 kernel: [75206.689992] CPU 0<br>       Sep 24 16:31:46 n1 kernel: [75206.690018] Modules linked in:<br>       ocfs2 crc32c libcrc32c nfsd auth_rpcgss exportfs ipmi_devintf<br>

       ipmi_si ipmi_msghandler ipv6 ocfs2_dlmfs ocfs2_dlm<br>       ocfs2_nodemanager configfs iptable_filter ip_tables x_tables<br>       xfs ib_iser rdma_cm ib_cm iw_cm ib_sa ib_mad ib_core ib_addr<br>       iscsi_tcp libiscsi scsi_transport_iscsi nfs lockd nfs_acl<br>

       sunrpc parport_pc lp parport loop serio_raw psmouse i2c_piix4<br>       i2c_core dcdbas evdev button k8temp shpchp pci_hotplug pcspkr<br>       ext3 jbd mbcache sg sr_mod cdrom sd_mod ata_generic pata_acpi<br>       usbhid hid ehci_hcd tg3 sata_svw pata_serverworks ohci_hcd<br>

       libata scsi_mod usbcore thermal processor fan fbcon tileblit<br>       font bitblit softcursor fuse<br>       Sep 24 16:31:46 n1 kernel: [75206.690455] Pid: 15931, comm:<br>       read_query Tainted: G      D 2.6.24-24-server #1<br>

       Sep 24 16:31:46 n1 kernel: [75206.690509] RIP:<br>       0010:[&lt;ffffffff8856c404&gt;]  [&lt;ffffffff8856c404&gt;]<br>       :ocfs2:ocfs2_meta_lock_full+0x6a4/0xec0<br>       Sep 24 16:31:46 n1 kernel: [75206.690591] RSP:<br>

       0018:ffff8101c64c9848  EFLAGS: 00010292<br>       Sep 24 16:31:46 n1 kernel: [75206.690623] RAX:<br>       0000000000000092 RBX: ffff81034ba74000 RCX: 00000000ffffffff<br>       Sep 24 16:31:46 n1 kernel: [75206.690659] RDX:<br>

       00000000ffffffff RSI: 0000000000000000 RDI: ffffffff8058ffa4<br>       Sep 24 16:31:46 n1 kernel: [75206.690695] RBP:<br>       0000000100080000 R08: 0000000000000000 R09: 00000000ffffffff<br>       Sep 24 16:31:46 n1 kernel: [75206.690730] R10:<br>

       0000000000000000 R11: 0000000000000000 R12: ffff81033fca4e00<br>       Sep 24 16:31:46 n1 kernel: [75206.690766] R13:<br>       ffff81033fca4f08 R14: ffff81033fca52b8 R15: ffff81033fca4f08<br>       Sep 24 16:31:46 n1 kernel: [75206.690802] FS:<br>

        00002b312f0119f0(0000) GS:ffffffff805c5000(0000)<br>       knlGS:00000000f546bb90<br>       Sep 24 16:31:46 n1 kernel: [75206.690857] CS:  0010 DS: 0000<br>       ES: 0000 CR0: 000000008005003b<br>       Sep 24 16:31:46 n1 kernel: [75206.690890] CR2:<br>

       00002b89f1e81000 CR3: 0000000168971000 CR4: 00000000000006e0<br>       Sep 24 16:31:46 n1 kernel: [75206.690925] DR0:<br>       0000000000000000 DR1: 0000000000000000 DR2: 0000000000000000<br>       Sep 24 16:31:46 n1 kernel: [75206.690961] DR3:<br>

       0000000000000000 DR6: 00000000ffff0ff0 DR7: 0000000000000400<br>       Sep 24 16:31:46 n1 kernel: [75206.690998] Process read_query<br>       (pid: 15931, threadinfo ffff8101c64c8000, task ffff81021543f7d0)<br>       Sep 24 16:31:46 n1 kernel: [75206.691054] Stack:<br>

        ffff810243c402af ffff810243c40299 ffff81021b462408<br>       000000011b462440<br>       Sep 24 16:31:46 n1 kernel: [75206.691116]  ffff8101c64c9910<br>       0000000100000000 ffff810217564e00 ffffffff8029018a<br>
       Sep 24 16:31:46 n1 kernel: [75206.691176]  0000000000000296<br>
       0000000000000001 ffffffffffffffff ffff81004c052f70<br>       Sep 24 16:31:46 n1 kernel: [75206.691217] Call Trace:<br>       Sep 24 16:31:46 n1 kernel: [75206.691273]<br>        [isolate_lru_pages+0x8a/0x210] isolate_lru_pages+0x8a/0x210<br>

       Sep 24 16:31:46 n1 kernel: [75206.691323]<br>        [&lt;ffffffff8857d4db&gt;] :ocfs2:ocfs2_delete_inode+0x16b/0x7e0<br>       Sep 24 16:31:46 n1 kernel: [75206.691362]<br>        [shrink_inactive_list+0x202/0x3c0]<br>

       shrink_inactive_list+0x202/0x3c0<br>       Sep 24 16:31:46 n1 kernel: [75206.691409]<br>        [&lt;ffffffff8857d370&gt;] :ocfs2:ocfs2_delete_inode+0x0/0x7e0<br>       Sep 24 16:31:46 n1 kernel: [75206.691449]<br>

        [fuse:generic_delete_inode+0xa8/0x450]<br>       generic_delete_inode+0xa8/0x140<br>       Sep 24 16:31:46 n1 kernel: [75206.691495]<br>        [&lt;ffffffff8857cd6d&gt;] :ocfs2:ocfs2_drop_inode+0x7d/0x160<br>       Sep 24 16:31:46 n1 kernel: [75206.691533]  [d_kill+0x3c/0x70]<br>

       d_kill+0x3c/0x70<br>       Sep 24 16:31:46 n1 kernel: [75206.691566]<br>        [prune_one_dentry+0xc1/0xe0] prune_one_dentry+0xc1/0xe0<br>       Sep 24 16:31:46 n1 kernel: [75206.691600]<br>        [prune_dcache+0x166/0x1c0] prune_dcache+0x166/0x1c0<br>

       Sep 24 16:31:46 n1 kernel: [75206.691635]<br>        [shrink_dcache_memory+0x3e/0x50] shrink_dcache_memory+0x3e/0x50<br>       Sep 24 16:31:46 n1 kernel: [75206.691670]<br>        [shrink_slab+0x124/0x180] shrink_slab+0x124/0x180<br>

       Sep 24 16:31:46 n1 kernel: [75206.691707]<br>        [try_to_free_pages+0x1e4/0x2f0] try_to_free_pages+0x1e4/0x2f0<br>       Sep 24 16:31:46 n1 kernel: [75206.691749]<br>        [__alloc_pages+0x196/0x3d0] __alloc_pages+0x196/0x3d0<br>

       Sep 24 16:31:46 n1 kernel: [75206.691790]<br>        [__do_page_cache_readahead+0xe0/0x210]<br>       __do_page_cache_readahead+0xe0/0x210<br>       Sep 24 16:31:46 n1 kernel: [75206.691834]<br>        [ondemand_readahead+0x117/0x1c0] ondemand_readahead+0x117/0x1c0<br>

       Sep 24 16:31:46 n1 kernel: [75206.691871]<br>        [do_generic_mapping_read+0x13d/0x3c0]<br>       do_generic_mapping_read+0x13d/0x3c0<br>       Sep 24 16:31:46 n1 kernel: [75206.691908]<br>        [file_read_actor+0x0/0x160] file_read_actor+0x0/0x160<br>

       Sep 24 16:31:46 n1 kernel: [75206.691949]<br>        [xfs:generic_file_aio_read+0xff/0x1b0]<br>       generic_file_aio_read+0xff/0x1b0<br>       Sep 24 16:31:46 n1 kernel: [75206.692026]<br>        [xfs:xfs_read+0x11c/0x250] :xfs:xfs_read+0x11c/0x250<br>

       Sep 24 16:31:46 n1 kernel: [75206.692067]<br>        [xfs:do_sync_read+0xd9/0xbb0] do_sync_read+0xd9/0x120<br>       Sep 24 16:31:46 n1 kernel: [75206.692101]<br>        [getname+0x1a9/0x220] getname+0x1a9/0x220<br>

       Sep 24 16:31:46 n1 kernel: [75206.692140]<br>        [&lt;ffffffff80254530&gt;] autoremove_wake_function+0x0/0x30<br>       Sep 24 16:31:46 n1 kernel: [75206.692185]<br>        [vfs_read+0xed/0x190] vfs_read+0xed/0x190<br>

       Sep 24 16:31:46 n1 kernel: [75206.692220]<br>        [sys_read+0x53/0x90] sys_read+0x53/0x90<br>       Sep 24 16:31:46 n1 kernel: [75206.692256]<br>        [system_call+0x7e/0x83] system_call+0x7e/0x83<br>       Sep 24 16:31:46 n1 kernel: [75206.692293]<br>

       Sep 24 16:31:46 n1 kernel: [75206.692316]<br>       Sep 24 16:31:46 n1 kernel: [75206.692317] Code: 0f 0b eb fe 83<br>       fd fe 0f 84 73 fc ff ff 81 fd 00 fe ff ff 0f<br>       Sep 24 16:31:46 n1 kernel: [75206.692483]  RSP &lt;ffff8101c64c9848&gt;<br>

<br><br>       Thanks<br>       Laurence<br><br>       _______________________________________________<br>       Ocfs2-users mailing list<br></div></div>       <a href="mailto:Ocfs2-users@oss.oracle.com" target="_blank">Ocfs2-users@oss.oracle.com</a> &lt;mailto:<a href="mailto:Ocfs2-users@oss.oracle.com" target="_blank">Ocfs2-users@oss.oracle.com</a>&gt; 
<div class="im"><br>       <a href="http://oss.oracle.com/mailman/listinfo/ocfs2-users" target="_blank">http://oss.oracle.com/mailman/listinfo/ocfs2-users</a><br>        <br><br><br></div></blockquote><br></blockquote></div>

<br></div>