Thank you for clarifying the overall deprecation of the replication feature 
rather than just some of the properties. This originated exploring blue green 
options for accumulo 2.x LTM.

"Another option might be to load separate Accumulo instances with an ETL 
pipeline using something like Apache Kafka or Apache NiFi. These alternatives 
tend to give you more control over where and how your data is loaded, and may 
result in a simpler and more reliable system architecture overall."

I'll look at these to that end.

I had poked around in zookeeper, but I'm unclear on where the call to the 
workers is persisting, having tried to unwind things.


Matthew M. Austin

Sr. DevSecOps Engineer

DSA, Inc.

8 Neshaminy Interplex Drive| Suite 209 | Trevose, PA 19053

www.dsainc.com<http://www.dsainc.com/> | 
[email protected]<mailto:[email protected]>



Culture makes a difference, learn about ours The DSA Way here: 
https://www.dsainc.com/the-dsa-way<https://url.emailprotection.link/?b8rNrM4MTPp9c5teJIa0ud060nquUWSdKmf-VkrYQbKNOmgfv-H3v1n4barDHp3uSr2rXffJyixmgWTH9KNbARg~~>





[cid:578917f7-8273-421e-ab01-233a4662f935]<http://www.dsainc.com/index.html>

Employee-Owned, Mission-Driven

Note: This email and any files transmitted with it are the property of Data 
Systems Analysts, Inc. and are privileged, confidential, and protected from 
disclosure. If the reader of this message is not the intended recipient, or an 
employee or agent responsible for delivering this message to the intended 
recipient, you are hereby notified that any dissemination, distribution, or 
copying of this communication is strictly prohibited. If you have received this 
communication in error, please notify the sender immediately by replying to the 
message and promptly delete it from your system.

________________________________
From: Christopher <[email protected]>
Sent: Friday, January 16, 2026 9:22 AM
To: [email protected] <[email protected]>
Subject: Re: accumulo 2.1.3 replication error help



________________________________
Hi Matthew,

While I don't know what is causing your specific problem, the cross-data center 
replication feature has had some outstanding bugs since its inception, and its 
lack of any active maintainers supporting that feature (along with its 
substantial complexity that entangled it to our write-ahead log implementation, 
which led to difficulty maintaining write-ahead log capability) led us to 
deprecate it as of 2.1.0, just over 4 years ago, and removed it in 3.0.0, 
though it may have already been broken prior to 2.1.0 (it's not a feature we 
get feedback on very often). You should be seeing warnings about the use of its 
properties in your logs, and compiler warnings using any replication-related 
APIs. The website properties documentation also describes the replication 
properties as deprecated.

While it may be possible that there is a simple fix to the issue you are 
seeing, I don't know what it is. I would strongly advise against attempting to 
use this deprecated feature. Since 2.1 is an LTM version, we are still 
maintaining 2.1 by publishing new releases, so we will still accept patches for 
2.1 to fix replication issues, if necessary to fix your issue, but to my 
knowledge, there are currently no active maintainers who have knowledge or 
interest in maintaining the feature.

Depending on your situation, there may be other viable alternatives. One option 
that some people may find helpful is the import/export table feature to perform 
incremental backups. Another option might be to load separate Accumulo 
instances with an ETL pipeline using something like Apache Kafka or Apache 
NiFi. These alternatives tend to give you more control over where and how your 
data is loaded, and may result in a simpler and more reliable system 
architecture overall.

It appears we didn't mention the deprecation in the 2.1.0 release notes or in 
the 2.x docs on the website, but did describe its removal, and reason for it, 
in 3.0.0's release notes.

I realize this is probably not the response you hoped for, but I hope that it 
is at least helpfully informative.

Kind regards,
Christopher

On Thu, Jan 15, 2026 at 11:30 PM Matthew Austin 
<[email protected]<mailto:[email protected]>> wrote:
Greetings,

I've got a development system and as I was working through setting up 
replication using steps outlined here
https://accumulo.apache.org/docs/2.x/administration/replication<https://accumulo.apache.org/docs/2.x/administration/replication>

I'm getting this error every 30 secs.

2026-01-16T04:25:15,960 [replication.WorkDriver] ERROR: Error while assigning 
work
java.lang.NullPointerException: null
at 
org.apache.accumulo.manager.replication.DistributedWorkQueueWorkAssigner.initializeWorkQueue(DistributedWorkQueueWorkAssigner.java:95)
 ~[accumulo-manager-2.1.3.jar:2.1.3]
at 
org.apache.accumulo.manager.replication.DistributedWorkQueueWorkAssigner.assignWork(DistributedWorkQueueWorkAssigner.java:107)
 ~[accumulo-manager-2.1.3.jar:2.1.3]
at org.apache.accumulo.manager.replication.WorkDriver.run(WorkDriver.java:85) 
~[accumulo-manager-2.1.3.jar:2.1.3]
at 
org.apache.accumulo.core.trace.TraceWrappedRunnable.run(TraceWrappedRunnable.java:52)
 ~[accumulo-core-2.1.3.jar:2.1.3]
at java.base/java.lang.Thread.run(Thread.java:829) [?:?]
2026-01-16T04:25:45,960 [replication.WorkDriver] ERROR: Error while assigning 
work
java.lang.NullPointerException: null
at 
org.apache.accumulo.manager.replication.DistributedWorkQueueWorkAssigner.initializeWorkQueue(DistributedWorkQueueWorkAssigner.java:95)
 ~[accumulo-manager-2.1.3.jar:2.1.3]
at 
org.apache.accumulo.manager.replication.DistributedWorkQueueWorkAssigner.assignWork(DistributedWorkQueueWorkAssigner.java:107)
 ~[accumulo-manager-2.1.3.jar:2.1.3]
at org.apache.accumulo.manager.replication.WorkDriver.run(WorkDriver.java:85) 
~[accumulo-manager-2.1.3.jar:2.1.3]
at 
org.apache.accumulo.core.trace.TraceWrappedRunnable.run(TraceWrappedRunnable.java:52)
 ~[accumulo-core-2.1.3.jar:2.1.3]
at java.base/java.lang.Thread.run(Thread.java:829) [?:?]
2026-01-16T04:26:15,960 [replication.WorkDriver] ERROR: Error while assigning 
work
java.lang.NullPointerException: null
at 
org.apache.accumulo.manager.replication.DistributedWorkQueueWorkAssigner.initializeWorkQueue(DistributedWorkQueueWorkAssigner.java:95)
 ~[accumulo-manager-2.1.3.jar:2.1.3]
at 
org.apache.accumulo.manager.replication.DistributedWorkQueueWorkAssigner.assignWork(DistributedWorkQueueWorkAssigner.java:107)
 ~[accumulo-manager-2.1.3.jar:2.1.3]
at org.apache.accumulo.manager.replication.WorkDriver.run(WorkDriver.java:85) 
~[accumulo-manager-2.1.3.jar:2.1.3]
at 
org.apache.accumulo.core.trace.TraceWrappedRunnable.run(TraceWrappedRunnable.java:52)
 ~[accumulo-core-2.1.3.jar:2.1.3]
at java.base/java.lang.Thread.run(Thread.java:829) [?:?]

I tried going in reverse and restarting things, but the error is still 
occurring. I have 2 tablets/hdfs datanodes, and 1 zookeeper for my little dev 
stack. Accumulo is otherwise functioning.

Thanks for any help.


Matthew M. Austin

Sr. DevSecOps Engineer

DSA, Inc.

8 Neshaminy Interplex Drive| Suite 209 | Trevose, PA 19053

www.dsainc.com<http://www.dsainc.com> | 
[email protected]<mailto:[email protected]>



Culture makes a difference, learn about ours The DSA Way here: 
https://www.dsainc.com/the-dsa-way<https://url.emailprotection.link/?b8rNrM4MTPp9c5teJIa0ud060nquUWSdKmf-VkrYQbKNOmgfv-H3v1n4barDHp3uSr2rXffJyixmgWTH9KNbARg~~>





[cid:ii_19bc7b58912defe94181]<http://www.dsainc.com/index.html>

Employee-Owned, Mission-Driven

Note: This email and any files transmitted with it are the property of Data 
Systems Analysts, Inc. and are privileged, confidential, and protected from 
disclosure. If the reader of this message is not the intended recipient, or an 
employee or agent responsible for delivering this message to the intended 
recipient, you are hereby notified that any dissemination, distribution, or 
copying of this communication is strictly prohibited. If you have received this 
communication in error, please notify the sender immediately by replying to the 
message and promptly delete it from your system.

Reply via email to