<html>
  <head>
    <meta content="text/html; charset=utf-8" http-equiv="Content-Type">
  </head>
  <body bgcolor="#FFFFFF" text="#000000">
    <br>
    <br>
    <div class="moz-cite-prefix">On 08/02/17 06:11, Nag Pavan Chilakam
      wrote:<br>
    </div>
    <blockquote
      cite="mid:1569385591.22125284.1486534262336.JavaMail.zimbra@redhat.com"
      type="cite">
      <pre wrap="">"gluster volume info" and "gluster vol status" would help in us debug faster.

However, coming to gfid mismatch, yes the file "abbreviations.log" (I assume the other brick copy also to be " abbreviations.log" and not "breviations.log" ....typo mistake?) is in gfid mismatch leading to IO error(gfid splitbrain)
Resolving data and metadata splitbrains are not recommended to be done from backend brick.
But in case of a GFID splitbrain(like in file abbreviations.log), the only method available is resolving from backend brick
You can read more about this in <a class="moz-txt-link-freetext" href="http://gluster.readthedocs.io/en/latest/Troubleshooting/split-brain/?highlight=gfid">http://gluster.readthedocs.io/en/latest/Troubleshooting/split-brain/?highlight=gfid</a>   (Fixing Directory entry split-brain   section)
(There is a bug already existing to resolve gfid splitbrain using CLI )

 </pre>
    </blockquote>
    I've read that doc, however I'm not sure what to do with bits that
    are not mentioned in that doc. Which is:<br>
    when some xattr does not exist on one copy but does on the other,
    like:<br>
    <br>
    3]$ getfattr -d -m . -e hex .vim.backup/.bash_profile.swp<br>
    # file: .vim.backup/.bash_profile.swp<br>
security.selinux=0x73797374656d5f753a6f626a6563745f723a756e6c6162656c65645f743a733000<br>
    trusted.afr.USER-HOME-client-0=0x000000010000000100000000<br>
    trusted.afr.USER-HOME-client-5=0x000000010000000100000000<br>
    <br>
    2]$ getfattr -d -m . -e hex .vim.backup/.bash_profile.swp<br>
    # file: .vim.backup/.bash_profile.swp<br>
security.selinux=0x73797374656d5f753a6f626a6563745f723a64656661756c745f743a733000<br>
    trusted.afr.USER-HOME-client-5=0x000000010000000100000000<br>
    trusted.afr.USER-HOME-client-6=0x000000010000000100000000<br>
    <br>
    unless the doc talks about it and I've gone (temporarily) blind, but
    if it's does not it would be great to include more scenarios/cases
    there.<br>
    many thx.<br>
    L.<br>
    <br>
    <blockquote
      cite="mid:1569385591.22125284.1486534262336.JavaMail.zimbra@redhat.com"
      type="cite">
      <pre wrap="">


thanks,
nagpavan


----- Original Message -----
From: "lejeczek" <a class="moz-txt-link-rfc2396E" href="mailto:peljasz@yahoo.co.uk">&lt;peljasz@yahoo.co.uk&gt;</a>
To: "Nag Pavan Chilakam" <a class="moz-txt-link-rfc2396E" href="mailto:nchilaka@redhat.com">&lt;nchilaka@redhat.com&gt;</a>
Cc: <a class="moz-txt-link-abbreviated" href="mailto:gluster-users@gluster.org">gluster-users@gluster.org</a>
Sent: Tuesday, 7 February, 2017 10:53:07 PM
Subject: Re: [Gluster-users] Input/output error - would not heal



On 07/02/17 12:50, Nag Pavan Chilakam wrote:
</pre>
      <blockquote type="cite">
        <pre wrap="">Hi,
Can you help us with more information on the volume, like volume status and volume info
One reason of "transport endpoint error" is the brick could be down

Also, i see that the syntax used for healing is wrong.
You need to use as below:
gluster v heal &lt;vname&gt; split-brain source-brick &lt;brick path&gt; &lt;filename considering brick path as /&gt;

In yourcase if brick path is "/G-store/1" and the file to be healed is "that_file" , then use below syntax (in this case i am considering "that_file" lying under the brick path directly"

gluster volume heal USER-HOME split-brain source-brick 10.5.6.100:/G-store/1 /that_file
</pre>
      </blockquote>
      <pre wrap="">
that was that, my copy-paste typo, it does not heal. 
Interestingly, that file is not reported by heal.

I've replied to -  GFID Mismatch - Automatic Correction ? - 
I think my problem is similar, here is a file the heal 
actually sees:


$ gluster vol heal USER-HOME info
Brick 
10.5.6.100:/__.aLocalStorages/3/0-GLUSTERs/0-USER.HOME/aUser/.vim.backup/.bash_profile.swp 

Status: Connected
Number of entries: 1

Brick 
10.5.6.49:/__.aLocalStorages/3/0-GLUSTERs/0-USER.HOME/aUser/.vim.backup/.bash_profile.swp 

Status: Connected
Number of entries: 1

I'm copying+pasting what I said in that reply to that thread:
...

yep, I'm seeing the same:
as follows:
3]$ getfattr -d -m . -e hex .
# file: .
security.selinux=0x73797374656d5f753a6f626a6563745f723a756e6c6162656c65645f743a733000
trusted.afr.USER-HOME-client-2=0x000000000000000000000000
trusted.afr.USER-HOME-client-3=0x000000000000000000000000
trusted.afr.USER-HOME-client-5=0x000000000000000000000000
trusted.afr.dirty=0x000000000000000000000000
trusted.gfid=0x06341b521ba94ab7938eca57f7a1824f
trusted.glusterfs.9e4ed9b7-373a-413b-bc82-b6f978e82ec4.xtime=0x5898e0cf000dd2fe
trusted.glusterfs.dht=0x000000010000000000000000ffffffff
trusted.glusterfs.quota.00000000-0000-0000-0000-000000000001.contri.1=0x00701c90fcb11200fffffef6f08c798e0000006a99819205
trusted.glusterfs.quota.dirty=0x3000
trusted.glusterfs.quota.size.1=0x00701c90fcb11200fffffef6f08c798e0000006a99819205
3]$ getfattr -d -m . -e hex .vim.backup
# file: .vim.backup
security.selinux=0x73797374656d5f753a6f626a6563745f723a756e6c6162656c65645f743a733000
trusted.afr.USER-HOME-client-3=0x000000000000000000000000
trusted.gfid=0x0b3a223955534de89086679a4dce8156
trusted.glusterfs.9e4ed9b7-373a-413b-bc82-b6f978e82ec4.xtime=0x5898621c0005d720
trusted.glusterfs.dht=0x000000010000000000000000ffffffff
trusted.glusterfs.quota.06341b52-1ba9-4ab7-938e-ca57f7a1824f.contri.1=0x000000000000040000000000000000020000000000000001
trusted.glusterfs.quota.dirty=0x3000
trusted.glusterfs.quota.size.1=0x000000000000040000000000000000020000000000000001
3]$ getfattr -d -m . -e hex .vim.backup/.bash_profile.swp
# file: .vim.backup/.bash_profile.swp
security.selinux=0x73797374656d5f753a6f626a6563745f723a756e6c6162656c65645f743a733000
trusted.afr.USER-HOME-client-0=0x000000010000000100000000
trusted.afr.USER-HOME-client-5=0x000000010000000100000000
trusted.gfid=0xc2693670fc6d4fed953f21dcb77a02cf
trusted.glusterfs.9e4ed9b7-373a-413b-bc82-b6f978e82ec4.xtime=0x5896043c000baa55
trusted.glusterfs.quota.0b3a2239-5553-4de8-9086-679a4dce8156.contri.1=0x00000000000000000000000000000001
trusted.pgfid.0b3a2239-5553-4de8-9086-679a4dce8156=0x00000001

2]$ getfattr -d -m . -e hex .
# file: .
security.selinux=0x73797374656d5f753a6f626a6563745f723a64656661756c745f743a733000
trusted.afr.USER-HOME-client-1=0x000000000000000000000000
trusted.afr.USER-HOME-client-2=0x000000000000000000000000
trusted.afr.USER-HOME-client-3=0x000000000000000000000000
trusted.afr.USER-HOME-client-5=0x000000000000000000000000
trusted.afr.dirty=0x000000000000000000000000
trusted.gfid=0x06341b521ba94ab7938eca57f7a1824f
trusted.glusterfs.9e4ed9b7-373a-413b-bc82-b6f978e82ec4.xtime=0x5898e0d000016f82
trusted.glusterfs.dht=0x000000010000000000000000ffffffff
trusted.glusterfs.quota.00000000-0000-0000-0000-000000000001.contri.1=0xa5e66200a7a45000cb96fbf7d6336229fae7152d8851097b
trusted.glusterfs.quota.dirty=0x3000
trusted.glusterfs.quota.size.1=0xa5e66200a7a45000cb96fbf7d6336229fae7152d8851097b
2]$ getfattr -d -m . -e hex .vim.backup
# file: .vim.backup
security.selinux=0x73797374656d5f753a6f626a6563745f723a64656661756c745f743a733000
trusted.afr.USER-HOME-client-3=0x000000000000000000000000
trusted.gfid=0x0b3a223955534de89086679a4dce8156
trusted.glusterfs.9e4ed9b7-373a-413b-bc82-b6f978e82ec4.xtime=0x5898621b000855fe
trusted.glusterfs.dht=0x000000010000000000000000ffffffff
trusted.glusterfs.quota.06341b52-1ba9-4ab7-938e-ca57f7a1824f.contri.1=0x000000000000040000000000000000020000000000000001
trusted.glusterfs.quota.dirty=0x3000
trusted.glusterfs.quota.size.1=0x000000000000040000000000000000020000000000000001
2]$ getfattr -d -m . -e hex .vim.backup/.bash_profile.swp
# file: .vim.backup/.bash_profile.swp
security.selinux=0x73797374656d5f753a6f626a6563745f723a64656661756c745f743a733000
trusted.afr.USER-HOME-client-5=0x000000010000000100000000
trusted.afr.USER-HOME-client-6=0x000000010000000100000000
trusted.gfid=0x8a5b6e4ad18a49d0bae920c9cf8673a5
trusted.glusterfs.9e4ed9b7-373a-413b-bc82-b6f978e82ec4.xtime=0x5896041400058191
trusted.glusterfs.quota.0b3a2239-5553-4de8-9086-679a4dce8156.contri.1=0x00000000000000000000000000000001
trusted.pgfid.0b3a2239-5553-4de8-9086-679a4dce8156=0x00000001


and the log bit:

GFID mismatch for 
&lt;gfid:335bf026-68bd-4bf4-9cba-63b65b12c0b1&gt;/abbreviations.xlsx 
6e9a7fa1-bfbe-4a59-ad06-a78ee1625649 on USER-HOME-client-6 
and 773b7ea3-31cf-4b24-94f0-0b61b573b082 on USER-HOME-client-0

most importantly, is there a workaround for the problem, as 
of now? Before the bug, it it's such, gets fixed.
b.w.
L.

-- end of paste

but I have a few more files which also report I/O errors and 
heal does NOT even mention them:
on the brick that is a "master"(samba was sharing to the users)

# file: abbreviations.log
security.selinux=0x73797374656d5f753a6f626a6563745f723a64656661756c745f743a733000
trusted.afr.dirty=0x000000000000000000000000
trusted.bit-rot.version=0x0200000000000000589081fd00060376
trusted.gfid=0x773b7ea331cf4b2494f00b61b573b082
trusted.glusterfs.quota.335bf026-68bd-4bf4-9cba-63b65b12c0b1.contri.1=0x0000000000002a000000000000000001
trusted.pgfid.335bf026-68bd-4bf4-9cba-63b65b12c0b1=0x00000001

on the "slave" brick, was not serving files (certainly not 
that file) to any users:

# file: bbreviations.log
security.selinux=0x73797374656d5f753a6f626a6563745f723a756e6c6162656c65645f743a733000
trusted.afr.dirty=0x000000000000000000000000
trusted.bit-rot.version=0x0200000000000000588c958a000b67ea
trusted.gfid=0x6e9a7fa1bfbe4a59ad06a78ee1625649
trusted.glusterfs.quota.335bf026-68bd-4bf4-9cba-63b65b12c0b1.contri.1=0x0000000000002a000000000000000001
trusted.pgfid.335bf026-68bd-4bf4-9cba-63b65b12c0b1=0x00000001

Question that probably was answered many times: is it OK to 
tamper with(remove in my case) files directly from bricks?
many thanks,
L.


</pre>
      <blockquote type="cite">
        <pre wrap="">regards,
nag pavan

----- Original Message -----
From: "lejeczek"<a class="moz-txt-link-rfc2396E" href="mailto:peljasz@yahoo.co.uk">&lt;peljasz@yahoo.co.uk&gt;</a>
<a class="moz-txt-link-abbreviated" href="mailto:To:gluster-users@gluster.org">To:gluster-users@gluster.org</a>
Sent: Tuesday, 7 February, 2017 2:00:51 AM
Subject: [Gluster-users] Input/output error - would not heal

hi all

I'm hitting such problem:

$ gluster vol heal USER-HOME split-brain source-brick
10.5.6.100:/G-store/1
Healing gfid:8a5b6e4a-d18a-49d0-bae9-20c9cf8673a5
failed:Transport endpoint is not connected.
Status: Connected
Number of healed entries: 0




$ gluster vol heal USER-HOME split-brain source-brick
10.5.6.100:/G-store/1/that_file
Lookup failed on /that_<a class="moz-txt-link-freetext" href="file:Input/output">file:Input/output</a>  error
Volume heal failed.

v3.9. it's a two-brick volume, was three but removed one I
think a few hours before the problem was first noticed.
what to do now?
many thanks,
L

_______________________________________________
Gluster-users mailing list
<a class="moz-txt-link-abbreviated" href="mailto:Gluster-users@gluster.org">Gluster-users@gluster.org</a>
<a class="moz-txt-link-freetext" href="http://lists.gluster.org/mailman/listinfo/gluster-users">http://lists.gluster.org/mailman/listinfo/gluster-users</a>
</pre>
      </blockquote>
      <pre wrap="">
</pre>
    </blockquote>
    <br>
  </body>
</html>