Recovering lost data » History » Version 3
Tom Clegg, 04/26/2019 07:58 PM
1 | 1 | Tom Clegg | h1. Untrashing lost blocks |
---|---|---|---|
2 | |||
3 | In some cases it is possible to recover data blocks that have been trashed by keep-balance (due to a bug or an install/config error). |
||
4 | |||
5 | If you suspect blocks have been trashed erroneously, you should immediately: |
||
6 | # On all keepstore servers: set EmptyTrashInterval to a long time like 2400h |
||
7 | # On all keepstore servers: restart keepstore |
||
8 | # Stop the keep-balance service |
||
9 | |||
10 | When you think you have corrected the underlying problem, you should: |
||
11 | # Set LostBlocksFile to a suitable value (perhaps "/tmp/keep-balance-lost-blocks.txt") in your keep-balance config |
||
12 | # Start keep-balance |
||
13 | |||
14 | After keep-balance completes its first sweep, inspect /tmp/keep-balance-lost-blocks.txt. If it's not empty, you can request all keepstores to untrash any blocks that are still recoverable with a script like this: |
||
15 | |||
16 | <pre><code class="bash"> |
||
17 | #!/bin/bash |
||
18 | set -e |
||
19 | |||
20 | # see Client.AuthToken in /etc/arvados/keep-balance/keep-balance.yml |
||
21 | token=xxxxxxx-your-system-auth-token-xxxxxxx |
||
22 | |||
23 | # all keep server hostnames |
||
24 | hosts=(keep0 keep1 keep2 keep3 keep4 keep5) |
||
25 | |||
26 | 3 | Tom Clegg | while read hash pdhs; do |
27 | 1 | Tom Clegg | echo "${hash}" |
28 | for h in ${hosts[@]}; do |
||
29 | 2 | Nico César | if curl -fgs -H "Authorization: Bearer $token" -X PUT "http://${h}:25107/untrash/$hash"; then |
30 | 1 | Tom Clegg | echo "${hash} ok ${host}" |
31 | fi |
||
32 | done |
||
33 | done < /tmp/keep-balance-lost-blocks.txt |
||
34 | </code></pre> |