WARNING A single process is both a transaction log and a storage server

[Vasilii (Vasiliy Popkov)]

WARNING: A single process is both a transaction log and a storage server.
For best performance use dedicated disks for the transaction logs by setting process classes.

We have this warning and have some question

we have
1 ssd for 2~3 process

what we must change

use 1 ssd for all process log
and other ssd for data

[fdbserver.4500]
datadir = /ssd1/data/foundationdb/data/$ID
logdir = /ssdLOG/log/foundationdb
[fdbserver.4501]
datadir = /ssd2/data/foundationdb/data/$ID
logdir = /ssdLOG/log/foundationdb

or
change for process data in one ssd and log other ssd for all process

[fdbserver.4500]
datadir = /ssd1/data/foundationdb/data/$ID
logdir = /ssd2/log/foundationdb
[fdbserver.4501]
datadir = /ssd2/data/foundationdb/data/$ID
logdir = /ssd3/log/foundationdb

now we use

[fdbserver.4500]
datadir = /ssd1/data/foundationdb/data/$ID
logdir = /ssd1/log/foundationdb
[fdbserver.4501]
datadir = /ssd2/data/foundationdb/data/$ID
logdir = /ssd2/log/foundationdb

Configuration:
Redundancy mode - triple
Storage engine - ssd-2
Coordinators - 5

Cluster:
FoundationDB processes - 300
Machines - 10
Memory availability - 8.5 GB per process on machine with least available
Retransmissions rate - 2 Hz
Fault Tolerance - 2 machines
Server time - 07/18/19 17:31:54

Data:
Replication health - Healthy
Moving data - 0.000 GB
Sum of key-value sizes - 692 MB
Disk space used - 35.854 GB

Operating space:
Storage server - 1696.9 GB free on most full server
Log server - 1696.9 GB free on most full server

Workload:
Read rate - 51 Hz
Write rate - 0 Hz
Transactions started - 12 Hz
Transactions committed - 0 Hz
Conflict rate - 0 Hz

Backup and DR:
Running backups - 0
Running DRs - 0

Client time: 07/18/19 17:31:35

WARNING: A single process is both a transaction log and a storage server.
For best performance use dedicated disks for the transaction logs by setting process classes.

[alexmiller]

@Vasilii

For best performance use dedicated disks for the transaction logs by setting process classes

Configuration — FoundationDB 7.1 is the official documentation on this subject, but there’s also been a number of threads over time. For example:

Cluster tuning cookbook

Hello everyone. I’m try to properly configure a cluster and struggling with the following behavior: There are 8 machines, each is 4 cores, 8GB of memory Each machine runs one FDB instance Engine is double ssd There are 5 coordinators in total Right now the transaction speed is just a little above 11k per second. Each transaction is 4 reads and 2 writes (non-atomics), key is around 64 bytes, value is less than 100 bytes. Looks like it’s quite underperforming and I’d appreciate some advice wh…

Production optimizations

Hi, Finally my cluster is up and running so I was able to do some significant testing. At this point, we’re using it as a cache, with a very simple transaction logic _, err = db.Transact(func(tr fdb.Transaction) (ret interface{}, err error) { future := tr.Get(sub.Pack(tuple.Tuple{"cache", ID})) res := future.MustGet() if len(res) == 0 { tr.Set(sub.Pack(tuple.Tuple{"cache", ID}), []byte(strconv.FormatInt(time.Now().UnixNano(), 10))) processChan <- id } ret…

Why doesn't my cluster performance scale when I double the number of machines?

Hello fellow FDB aficionado! I was wondering if you could help me identifying why I am not seeing any improvements after I scale my FDB cluster out. I have read the performance page and I know it is supposed to scale… I must be doing something incorrectly. Explanation: Let me be more specific with a benchmark I just run today using FDB 5.2 on linux on AWS/EC2. I started a cluster of 10 instances (i3.xlarge with NVME SSD 900GB high network throughput, 4vCPU). For information I ha…

[Rjerk]

Hi @alexmiller

If a process has class type ‘log’ but it has both a transaction log and a storage server.

How to detach the storage role from the process?

I already tried to restart the process but it’s not working.