[
https://issues.apache.org/jira/browse/ZOOKEEPER-4334?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=17389792#comment-17389792
]
Damien Diederen commented on ZOOKEEPER-4334:
--------------------------------------------
[~ekleszcz] wrote:
bq. that won't solve the problem as the change considers only the SASL auth
between the quorum members and my case regards the Java client to server auth.
Ah, right; I just spotted "the quorum member's saslToken is null," saw that you
were using keytabs, assumed this was about quorum auth, and thought I'd mention
ZOOKEEPER-4030.
bq. I have just discovered the extra flag:
{{zookeeper.sasl.client.canonicalize.hostname}}. This means that by default we
have to strictly use the canonical names for the principals. What I would like
to achieve instead is to define the aliases in the principals. \[…\] Tested and
it keeps failing \[…\]
Right. As [~eolivelli] mentions, Kerberos implementations tend to be bound to
"real" names, as returned by reverse DNS resolution.
ZooKeeper \(client-to-server, and now server-to-server) supports referencing
members using aliases, but the correct tickets still have to be provided.
My understanding is that this is a Kerberos limitation, not a ZooKeeper issue.
You are of course welcome to suggest a workaround if you find one, but I would
otherwise suggest amending or closing this ticket.
> SASL authentication fails when using host aliases
> -------------------------------------------------
>
> Key: ZOOKEEPER-4334
> URL: https://issues.apache.org/jira/browse/ZOOKEEPER-4334
> Project: ZooKeeper
> Issue Type: Bug
> Affects Versions: 3.6.1
> Reporter: Emil Kleszcz
> Priority: Critical
>
> I faced an issue while trying to use alternative aliases with Zookeeper
> quorum when SASL is enabled. The errors I get in zookeeper log are the
> following:
> ```
> 2021-07-12 21:04:46,437 [myid:3] - WARN
> [NIOWorkerThread-3:ZooKeeperServer@1661] - Client /<IP addr>:37368 failed to
> SASL authenticate: {}
> javax.security.sasl.SaslException: GSS initiate failed [Caused by
> GSSException: Failure unspecified at GSS-API level (Mechanism level: Checksum
> failed)]
> at
> com.sun.security.sasl.gsskerb.GssKrb5Server.evaluateResponse(GssKrb5Server.java:199)
> at
> org.apache.zookeeper.server.ZooKeeperSaslServer.evaluateResponse(ZooKeeperSaslServer.java:49)
> at
> org.apache.zookeeper.server.ZooKeeperServer.processSasl(ZooKeeperServer.java:1650)
> at
> org.apache.zookeeper.server.ZooKeeperServer.processPacket(ZooKeeperServer.java:1599)
> at
> org.apache.zookeeper.server.NIOServerCnxn.readRequest(NIOServerCnxn.java:379)
> at
> org.apache.zookeeper.server.NIOServerCnxn.readPayload(NIOServerCnxn.java:182)
> at org.apache.zookeeper.server.NIOServerCnxn.doIO(NIOServerCnxn.java:339)
> at
> org.apache.zookeeper.server.NIOServerCnxnFactory$IOWorkRequest.doWork(NIOServerCnxnFactory.java:522)
> at
> org.apache.zookeeper.server.WorkerService$ScheduledWorkRequest.run(WorkerService.java:154)
> at
> java.util.concurrent.ThreadPoolExecutor.runWorker(ThreadPoolExecutor.java:1149)
> at
> java.util.concurrent.ThreadPoolExecutor$Worker.run(ThreadPoolExecutor.java:624)
> at java.lang.Thread.run(Thread.java:748)
> Caused by: GSSException: Failure unspecified at GSS-API level (Mechanism
> level: Checksum failed)
> at sun.security.jgss.krb5.Krb5Context.acceptSecContext(Krb5Context.java:856)
> at sun.security.jgss.GSSContextImpl.acceptSecContext(GSSContextImpl.java:342)
> at sun.security.jgss.GSSContextImpl.acceptSecContext(GSSContextImpl.java:285)
> at
> com.sun.security.sasl.gsskerb.GssKrb5Server.evaluateResponse(GssKrb5Server.java:167)
> ... 11 more
> Caused by: KrbException: Checksum failed
> at
> sun.security.krb5.internal.crypto.Aes256CtsHmacSha1EType.decrypt(Aes256CtsHmacSha1EType.java:102)
> at
> sun.security.krb5.internal.crypto.Aes256CtsHmacSha1EType.decrypt(Aes256CtsHmacSha1EType.java:94)
> at sun.security.krb5.EncryptedData.decrypt(EncryptedData.java:175)
> at sun.security.krb5.KrbApReq.authenticate(KrbApReq.java:281)
> at sun.security.krb5.KrbApReq.<init>(KrbApReq.java:149)
> at
> sun.security.jgss.krb5.InitSecContextToken.<init>(InitSecContextToken.java:108)
> at sun.security.jgss.krb5.Krb5Context.acceptSecContext(Krb5Context.java:829)
> ... 14 more
> Caused by: java.security.GeneralSecurityException: Checksum failed
> at
> sun.security.krb5.internal.crypto.dk.AesDkCrypto.decryptCTS(AesDkCrypto.java:451)
> at
> sun.security.krb5.internal.crypto.dk.AesDkCrypto.decrypt(AesDkCrypto.java:272)
> at sun.security.krb5.internal.crypto.Aes256.decrypt(Aes256.java:76)
> at
> sun.security.krb5.internal.crypto.Aes256CtsHmacSha1EType.decrypt(Aes256CtsHmacSha1EType.java:100)
> ... 20 more
> ```
> What did I do?
> 1) created host aliases for each quorum node (a,b,c): zk1, zk2, zk3
> 2) Changed in zoo.cfg:
> changed from
> server.1=a
> server.2=b
> server.3=c
> to:
> server.1=zk1
> server.2=zk2
> server.3=zk3
> (at this stage after restarting the ensemble all works as expected.
> 3) Generate new keytab with alias-based principals and host-based principals
> in zookeeper.keytab
> 4) Change jaas.conf (server) definition from:
> Server
> { com.sun.security.auth.module.Krb5LoginModule required useKeyTab=true
> keyTab="/etc/zookeeper/conf/zookeeper.keytab" storeKey=true
> useTicketCache=false principal="zookeeper/a.com@COM"; }
> ;
> to
> Server
> { com.sun.security.auth.module.Krb5LoginModule required useKeyTab=true
> keyTab="/etc/zookeeper/conf/zookeeper.keytab" storeKey=true
> useTicketCache=false principal="zookeeper/zk1.com@COM"; }
> ;
> From that moment, after restarting quorum members, I get the above error.
> Now, why do I do this?
> To allow other services such as zkfc,hbase,hdfs,yarn to connect to the
> quorum using aliases. Interestingly, without changing the zookeeper
> principal, hbase works perfectly, but the other 3 services fail with:
> ```
> <2021-07-12T20:45:19.491+0200> <INFO> <org.apache.zookeeper.ZooKeeper>:
> <Initiating client connection,
> connectString=zk01.com:2181,zk02.com:2181,zk03.com:2181 sessionTimeout=10000
> watcher=org.apache.hadoop.ha.ActiveStandbyElector$WatcherWithClientRef@3246fb96>
> <2021-07-12T20:45:19.519+0200> <INFO> <org.apache.zookeeper.Login>: <Client
> successfully logged in.>
> <2021-07-12T20:45:19.521+0200> <INFO> <org.apache.zookeeper.Login>: <TGT
> refresh thread started.>
> <2021-07-12T20:45:19.524+0200> <INFO> <org.apache.zookeeper.Login>: <TGT
> valid starting at: Mon Jul 12 20:45:19 CEST 2021>
> <2021-07-12T20:45:19.524+0200> <INFO> <org.apache.zookeeper.Login>: <TGT
> expires: Tue Jul 13 21:45:19 CEST 2021>
> <2021-07-12T20:45:19.524+0200> <INFO> <org.apache.zookeeper.Login>: <TGT
> refresh sleeping until: Tue Jul 13 17:05:16 CEST 2021>
> <2021-07-12T20:45:19.524+0200> <INFO>
> <org.apache.zookeeper.client.ZooKeeperSaslClient>: <Client will use GSSAPI as
> SASL mechanism.>
> <2021-07-12T20:45:19.530+0200> <INFO> <org.apache.zookeeper.ClientCnxn>:
> <Opening socket connection to server zk02.com/<ip addr>:2181. Will attempt to
> SASL-authenticate using Login Context section 'Client'>
> <2021-07-12T20:45:19.535+0200> <INFO> <org.apache.zookeeper.ClientCnxn>:
> <Socket connection established to zk02.com/<ip addr>:2181, initiating session>
> <2021-07-12T20:45:19.543+0200> <INFO> <org.apache.zookeeper.ClientCnxn>:
> <Session establishment complete on server zk02.com/<ip addr>:2181, sessionid
> = 0x200247870fb0007, negotiated timeout = 10000>
> <2021-07-12T20:45:19.561+0200> <ERROR>
> <org.apache.zookeeper.client.ZooKeeperSaslClient>: <SASL authentication
> failed using login context 'Client' with exception: {}>
> javax.security.sasl.SaslException: Error in authenticating with a Zookeeper
> Quorum member: the quorum member's saslToken is null.
> at
> org.apache.zookeeper.client.ZooKeeperSaslClient.createSaslToken(ZooKeeperSaslClient.java:279)
> at
> org.apache.zookeeper.client.ZooKeeperSaslClient.respondToServer(ZooKeeperSaslClient.java:242)
> at
> org.apache.zookeeper.ClientCnxn$SendThread.readResponse(ClientCnxn.java:805)
> at org.apache.zookeeper.ClientCnxnSocketNIO.doIO(ClientCnxnSocketNIO.java:94)
> at
> org.apache.zookeeper.ClientCnxnSocketNIO.doTransport(ClientCnxnSocketNIO.java:366)
> at org.apache.zookeeper.ClientCnxn$SendThread.run(ClientCnxn.java:1145)
> <2021-07-12T20:45:19.564+0200> <INFO> <org.apache.zookeeper.ClientCnxn>:
> <Unable to read additional data from server sessionid 0x200247870fb0007,
> likely server has closed socket, closing socket connection and attempting
> reconnect>
> <2021-07-12T20:45:19.671+0200> <INFO>
> <org.apache.hadoop.ha.ActiveStandbyElector>: <Session connected.>
> <2021-07-12T20:45:19.672+0200> <ERROR>
> <org.apache.hadoop.hdfs.tools.DFSZKFailoverController>:
> <DFSZKFailOverController exiting due to earlier exception
> java.io.IOException: Couldn't determine existence of znode
> ```
> When I change the principle of zookeeper hbase starts failing with this
> error and other services except for the zookeeper itself is somehow working
> fine. After that, I cannot connect manually to the zk quorum using zkCli and
> zookeeper-client with all possible combinations of principals.
> I wonder if that may have something to do with the "Server
> environment:host.name=" pointing to the canonical name (and not the alias)
> during the startup. The same happens after specifying the alias with
> clientPortAddress=.
--
This message was sent by Atlassian Jira
(v8.3.4#803005)
