Not enough free bitmap slots when setting up a 3 node DRBD

DRBD
1

I can’t figure out why I am getting a failure on bringing up my resource. It says there are not enough bitmap slots, but I have only 12 resources on there and I even added more slots.

:~$ sudo drbdadm up RT6
15: Failure: (162) Invalid configuration request
additional info from kernel:
Not enough free bitmap slots (available=0, needed=1)

Command ‘drbdsetup attach 15 /dev/vg0/RT6 /dev/vg0/RT6 internal’ terminated with exit code 10
:~$ cat /sys/module/drbd/parameters/minor_count
128

128 seems excessive…

I was trying to add a third node. All my 2 note sets are fine.

resource RT6 {
device /dev/drbd15;
disk /dev/vg0/RT6;
meta-disk internal;
on trk-kvm-01 {
node-id 0;
}
on trk-kvm-02 {
node-id 1;
}
on trk-kvm-04 {
node-id 2;
}
connection {
host trk-kvm-01 address 10.10.1.145:7115;
host trk-kvm-02 address 10.10.1.146:7116;
}
connection {
host trk-kvm-01 address 10.10.3.15:7315;
host trk-kvm-04 address 10.10.3.17:7357;
}
connection {
host trk-kvm-02 address 10.10.2.16:7256;
host trk-kvm-04 address 10.10.2.187:7257;
}
}

Is my config wrong? Why might I be getting that error? I saw this about not reading bitmaps

and did update to 9.3.1/9.34.8-0:

:~$ sudo drbdadm --version
DRBDADM_BUILDTAG=GIT-hash:\ 696ee0fafa437f3b33b1e437cfc2741bf93604b2\ build\ by\ buildd@lcy02-amd64-073,\ 2026-03-10\ 12:39:25
DRBDADM_API_VERSION=2
DRBD_KERNEL_VERSION_CODE=0x090301
DRBD_KERNEL_VERSION=9.3.1
DRBDADM_VERSION_CODE=0x092200
DRBDADM_VERSION=9.34.0-rc.1

But I am not sure it is related.

It would seem to me the message is a bit of a red herring, that there should be slots available.

What am I missing?

2

I find that even if I try this now using two nodes only

resource RT6 {
device /dev/drbd15;
disk  /dev/vg0/RT6;
meta-disk internal;
on trk-kvm-01 {
node-id   0;
}
#on trk-kvm-02 {
#node-id   1;
#}
on trk-kvm-04 {
node-id   2;
}
#connection {
#host  trk-kvm-01  address 10.10.1.145:7115;
#host  trk-kvm-02  address 10.10.1.146:7116;
#}
connection {
host  trk-kvm-01  address 10.10.3.15:7315;
host  trk-kvm-04  address 10.10.3.17:7357;
}
#connection {
#host  trk-kvm-02  address 10.10.2.16:7256;
#host  trk-kvm-04  address 10.10.2.187:7257;
#}
}

I still get the message:

4:~$ sudo drbdadm up RT6
15: Failure: (162) Invalid configuration request
additional info from kernel:
Not enough free bitmap slots (available=0, needed=1)

Command ‘drbdsetup attach 15 /dev/vg0/RT6 /dev/vg0/RT6 internal’ terminated with exit code 10

I am at a loss to understand/explain this. It can’t be a simple resource count, I have only a few. I have run many more before.

3

Best guess is that there was something weird which happened during the initial meta-data creation. Was this perhaps originally created as a one node setup?

The create-md command should parse the configuration and create the needed amount of bitmap slots in the DRBD meta-data. If you want to specify the number of meta-data slots manually you could also do that via the --max-peers= option.

At this point I would advise you readjust the configuration for the three nodes you need and then re-run the drbdadm create-md RT6 command on all nodes and try again.

1 Like
4

That is a thing of beauty.

I was originally a 3 node, I made it two node for a while and was returning it to 3 node so it never occurred to me to recreate the metadata. In hindsight, now you bring it to my attention I may have recreated the metadata as two node at some point, it is quite possible.

Anyway, I am syncing now and I know now what to do should I encounter this again. This is VERY much appreciated.

1 Like