zookeeper-dev mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From "Jiafu Jiang (JIRA)" <j...@apache.org>
Subject [jira] [Created] (ZOOKEEPER-3220) Snapshot is not written to disk and cause data inconsistency.
Date Tue, 18 Dec 2018 09:56:00 GMT
Jiafu Jiang created ZOOKEEPER-3220:
--------------------------------------

             Summary: Snapshot is not written to disk and cause data inconsistency.
                 Key: ZOOKEEPER-3220
                 URL: https://issues.apache.org/jira/browse/ZOOKEEPER-3220
             Project: ZooKeeper
          Issue Type: Bug
          Components: server
    Affects Versions: 3.4.13, 3.4.12
            Reporter: Jiafu Jiang


We known that ZooKeeper server will call fsync to make sure that log data has been successfully
saved to disk. But ZooKeeper server does not call fsync to make sure that a snapshot has been successfully
saved, which may cause potential problems. Since a close to a file description does not
make sure that data is written to disk, see [http://man7.org/linux/man-pages/man2/close.2.html#notes] for
more details.

 

If the snapshot is not successfully  saved to disk, it may lead to data inconsistency.
Here is my example, which is also a real problem I have ever met.

1. I deployed a 3-node ZooKeeper cluster: zk1, zk2, and zk3, zk2 was the leader.

2. Both zk1 and zk2 had the log records from log1~logX, X is the zxid.

3. The machine of zk1 restarted, and during the reboot,  log(X+1) ~ log Y are saved to log
files of both zk2(leader) and zk3(follower).

4. After zk1 restarted successfully, it found itself to be a follower, and it began to synchronize
log with the leader. The leader sent a snapshot(records from log 1 ~ log Y) to zk1, zk1 saved
the snapshot to local disk by calling the method ZooKeeperServer.takeSnapshot. But unfortunately,
when the method returned, the snapshot data was not saved to disk yet. If fact the snapshot
file was created, but the size was 0.

5. zk1 finished the synchronization and began to accept new request from the leader. Say log(Y
+ 1) ~ log Z was accepted by zk1 and  saved to log file. With fsync zk1 can make sure log
data is not lost.

6. zk1 restarted again. Since the snapshot's size was 0, it would not be used, therefore zk1
recovered using the log files. But the records from log(X+1) ~ logY were lost ! 

 

Sorry for my poor English.

 



--
This message was sent by Atlassian JIRA
(v7.6.3#76005)

Mime
View raw message