### Author:  [@losfair](https://forums.foundationdb.org/u/losfair)
#### Post date: [April 25, 2023, 10:38am UTC](https://forums.foundationdb.org/t/100-cpu-usage-on-an-idle-cluster/3928/1 "2023-04-25T10:38:42Z")
Sometimes we are seeing 100% CPU usage on DR destination clusters and a high rate of “transaction started”, even when no client is connected and `dr_agent` has been stopped:
```auto
fdb> status details
Using cluster file `/etc/foundationdb/fdb.cluster'.
Configuration:
Storage engine - ssd-2
Coordinators - 1
Exclusions - 2 (type `exclude' for details)
Usable Regions - 1
Cluster:
FoundationDB processes - 1
Zones - 1
Machines - 1
Memory availability - 7.3 GB per process on machine with least available
Fault Tolerance - 0 machines
Server time - 04/25/23 10:31:29
Data:
Replication health - Healthy
Moving data - 0.000 GB
Sum of key-value sizes - 20 MB
Disk space used - 238 MB
Operating space:
Storage server - 46.0 GB free on most full server
Log server - 46.0 GB free on most full server
Workload:
Read rate - 9 Hz
Write rate - 0 Hz
Transactions started - 35190 Hz
Transactions committed - 0 Hz
Conflict rate - 0 Hz
Backup and DR:
Running backups - 0
Running DRs - 1 as secondary
Running DR tags (as secondary):
default -
Process performance details:
10.220.11.2:4500:tls (100% cpu; 51% machine; 0.000 Gbps; 0% disk IO; 3.1 GB / 7.3 GB RAM )
Coordination servers:
10.220.11.2:4500:tls (reachable)
Client time: 04/25/23 10:31:25
```
Restarting the cluster by running `kill all` recovers it out of the bad state. The issue is hard to reproduce, since we only see it happen about once per cluster per week randomly.
Logs are mostly of type `TransactionMetrics`, and sometimes `GlobalConfig_RetryableMigrationError`:
```xml
```
The nodes are running FoundationDB 7.1.31.
---
### Author:  [@mengxu](https://forums.foundationdb.org/u/mengxu)
#### Post date: [April 26, 2023, 12:47am UTC](https://forums.foundationdb.org/t/100-cpu-usage-on-an-idle-cluster/3928/2 "2023-04-26T00:47:44Z")
after 6.3, FDB is not maintaining the old-style DR (which has two FDB clusters connected by dr\_agent).
FDB uses HA to replace DR, which uses less resource and provide no-data loss guarantee in the failover
---
### Author:  [@losfair](https://forums.foundationdb.org/u/losfair)
#### Post date: [April 26, 2023, 2:11am UTC](https://forums.foundationdb.org/t/100-cpu-usage-on-an-idle-cluster/3928/3 "2023-04-26T02:11:50Z")
Is it possible to have more than two regions with the new HA configuration?
Thanks for the reply! @mengxu
---
### Author:  [@mengxu](https://forums.foundationdb.org/u/mengxu)
#### Post date: [April 26, 2023, 3:58am UTC](https://forums.foundationdb.org/t/100-cpu-usage-on-an-idle-cluster/3928/4 "2023-04-26T03:58:16Z")
Nope. only 2 regions supported so far.
Region 1 has 1 primary dc and 1 satellite DC; and
Region 2 has 1 dc and 1 satellite DC.
---
### Author:  [@losfair](https://forums.foundationdb.org/u/losfair)
#### Post date: [May 2, 2023, 2:43pm UTC](https://forums.foundationdb.org/t/100-cpu-usage-on-an-idle-cluster/3928/5 "2023-05-02T14:43:33Z")
We have a fix for the issue with DR itself, but it seems that the FDB repository is closed for pull requests. Would you be willing to review and accept a patch? @mengxu
>
---
### Author:  [@mengxu](https://forums.foundationdb.org/u/mengxu)
#### Post date: [May 2, 2023, 4:38pm UTC](https://forums.foundationdb.org/t/100-cpu-usage-on-an-idle-cluster/3928/6 "2023-05-02T16:38:45Z")
@ammolitor is managing the permission. He can do that.
---
### Author:  [@losfair](https://forums.foundationdb.org/u/losfair)
#### Post date: [May 3, 2023, 3:02pm UTC](https://forums.foundationdb.org/t/100-cpu-usage-on-an-idle-cluster/3928/7 "2023-05-03T15:02:13Z")
Thanks!
Hi @ammolitor , can I get permission to open pull requests to the foundationdb repository?
---
### Author:  [@ammolitor](https://forums.foundationdb.org/u/ammolitor)
#### Post date: [May 3, 2023, 3:15pm UTC](https://forums.foundationdb.org/t/100-cpu-usage-on-an-idle-cluster/3928/8 "2023-05-03T15:15:57Z")
invite sent, thanks for contributing!
---
### Author:  [@losfair](https://forums.foundationdb.org/u/losfair)
#### Post date: [May 3, 2023, 3:42pm UTC](https://forums.foundationdb.org/t/100-cpu-usage-on-an-idle-cluster/3928/9 "2023-05-03T15:42:50Z")
Didn’t receive the email, where should I check for the invite?
My GitHub username is `losfair`
---
### Author:  [@ammolitor](https://forums.foundationdb.org/u/ammolitor)
#### Post date: [May 3, 2023, 3:55pm UTC](https://forums.foundationdb.org/t/100-cpu-usage-on-an-idle-cluster/3928/10 "2023-05-03T15:55:37Z")
> [@losfair](#):
>
> losfair
my mistake, I assumed you were the owner of the fork where the referenced commit was shared from. you should have it now.