You are viewing a plain text version of this content. The canonical link for it is here.
Posted to issues@zookeeper.apache.org by "Antoine Tran (Jira)" <ji...@apache.org> on 2020/08/12 13:35:00 UTC
[jira] [Commented] (ZOOKEEPER-3466) ZK cluster converges, but does
not properly handle client connections (new in 3.5.5)
[ https://issues.apache.org/jira/browse/ZOOKEEPER-3466?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=17176339#comment-17176339 ]
Antoine Tran commented on ZOOKEEPER-3466:
-----------------------------------------
Hi, we have the exact same symptoms. We can reproduce a few time but only after an hour or so. Our env:
* physical machines or clouds
* redhat 7.7
* Rancher 2.4.3 RKE kubernetes 1.17.5
* zookeeper 3.6.1 deployed as static HA cluster with 3 nodes, as Kubernetes statefulset
Everything works until sometime, then it looks like the leader is unable to work, and the followers can't do anything.
Some logs:
{code:java}
[rancher@dl2psto-s00 infrastructure]$ kl2 exec -it zookeeper-cl-0 -- sh -c "echo srvr | nc localhost 2181"
Zookeeper version: 3.6.1--104dcb3e3fb464b30c5186d229e00af9f332524b, built on 04/21/2020 15:01 GMT
Latency min/avg/max: 0/0.3019/12
Received: 15962
Sent: 7162
Connections: 11
Outstanding: 8799
Zxid: 0x200000000
Mode: leader
Node count: 256718
Proposal sizes last/min/max: 48/36/41704
[rancher@dl2psto-s00 infrastructure]$ kl2 exec -it zookeeper-cl-1 -- sh -c "echo srvr | nc localhost 2181"
Zookeeper version: 3.6.1--104dcb3e3fb464b30c5186d229e00af9f332524b, built on 04/21/2020 15:01 GMT
Latency min/avg/max: 0/0.6582/9
Received: 9125
Sent: 620
Connections: 6
Outstanding: 8416
Zxid: 0x10037dee1
Mode: follower
Node count: 256718
[rancher@dl2psto-s00 infrastructure]$ kl2 exec -it zookeeper-cl-2 -- sh -c "echo srvr | nc localhost 2181"
Zookeeper version: 3.6.1--104dcb3e3fb464b30c5186d229e00af9f332524b, built on 04/21/2020 15:01 GMT
Latency min/avg/max: 0/0.25/9
Received: 17539
Sent: 9039
Connections: 11
Outstanding: 8406
Zxid: 0x10037dee1
Mode: follower
Node count: 256718{code}
I notice the leader (always) have
{code:java}
Zxid: 0x200000000{code}
Tcp is OK but
{code:java}
zkCli.sh ls /{code}
responds in timeout whatever the server.
Zookeeper configuration:
{code:java}
syncLimit=5
dataDir=/space/ZookeeperData/datadir
clientPort=2181
maxClientCnxns=6000
autopurge.snapRetainCount=3
autopurge.purgeInterval=1
server.1=zookeeper-cl-0.zk-leader:2888:3888
server.2=0.0.0.0:2888:3888
server.3=zookeeper-cl-2.zk-leader:2888:3888
{code}
One particularity of our deployment is Zookeeper datadir is in a tmpfs. Then we thought we reached some hard limit (memory hard limit of 6GB) but our memory consumption seems ok:
{code:java}
[rancher@dl2psto-s00 infrastructure]$ kl2 exec -it zookeeper-cl-1 -- df -h | grep datadir
tmpfs 126G 585M 126G 1% /space/ZookeeperData/datadir
docker stats (3 nodes): ~687.3MiB / 6GiB
{code}
> ZK cluster converges, but does not properly handle client connections (new in 3.5.5)
> ------------------------------------------------------------------------------------
>
> Key: ZOOKEEPER-3466
> URL: https://issues.apache.org/jira/browse/ZOOKEEPER-3466
> Project: ZooKeeper
> Issue Type: Bug
> Affects Versions: 3.5.5
> Environment: Linux
> Reporter: Jan-Philip Gehrcke
> Priority: Major
>
> Hey, we explore switching from ZooKeeper 3.4.14 to ZooKeeper 3.5.5 in [https://github.com/dcos/dcos].
> DC/OS coordinates ZooKeeper via Exhibitor. We are not changing anything w.r.t. Exhibitor for now, and are hoping that we can use ZooKeeper 3.5.5 as a drop-in replacement for 3.4.14. This seems to work fine when Exhibitor uses a so-called static ensemble where the individual ZooKeeper instances are known a priori.
> When Exhibitor however discovers individual ZooKeeper instances ("dynamic" back-end) then I think we observe a regression where ZooKeeper 3.5.5 can get into the following bad state (often, but not always):
> # three ZooKeeper instances find each other, leader election takes place (*expected*)
> # leader election succeeds: two followers, one leader (*expected*)
> # all three ZK instances respond IAMOK to RUOK (*expected*)
> # all three ZK instances respond to SRVR (one says "Mode: leader", the other two say "Mode: follower") (*expected*)
> # all three ZK instances respond to MNTR and show plausible output (*expected*)
> # *{color:#ff0000}Unexpected:{color}* any ZooKeeper client trying to connect to any of the three nodes observes a "connection timeout", whereas notably this is *not* a TCP connect() timeout. The TCP connect() succeeds, but then ZK does not seem to send the expected byte sequence to the TCP connection, and the ZK client waits for it via recv() until it hits a timeout condition. Examples for two different clients:
> ## In Kazoo we specifically hit _Connection time-out: socket time-out during read_
> generated here: [https://github.com/python-zk/kazoo/blob/88b657a0977161f3815657878ba48f82a97a3846/kazoo/protocol/connection.py#L249]
> ## In zkCli we see _Client session timed out, have not heard from server in 15003ms for sessionid 0x0, closing socket connection and attempting reconnect (org.apache.zookeeper.ClientCnxn:main-SendThread(localhost:2181))_
> # This state is stable, it will last forever (well, at least for multiple hours and we didn't test longer than that).
> # In our system the ZooKeeper clients are crash-looping. They retry. What I have observed is that while they retry the ZK ensemble accumulates outstanding requests, here shown from MNTR output (emphasis mine):
> zk_packets_received 2008
> zk_packets_sent 127
> zk_num_alive_connections 18
> zk_outstanding_requests *1880*
> # The leader emits log lines confirming session timeout, example:
> _[myid:3] INFO [SessionTracker:ZooKeeperServer@398] - Expiring session 0x2000642b18f0020, timeout of 10000ms exceeded [myid:3] INFO [SessionTracker:QuorumZooKeeperServer@157] - Submitting global closeSession request for session 0x2000642b18f0020_
> # In this state, restarting any one of the two ZK followers results in the same state (clients don't get data from ZK upon connect).
> # In this state, restarting the ZK leader, and therefore triggering a leader re-election, almost immediately results in all clients being able to connect to all ZK instances successfully.
--
This message was sent by Atlassian Jira
(v8.3.4#803005)