- What we know
- What we've created
- Hints and Kinks
- Checking Corosync cluster membership
- Configuring radosgw to behave like Amazon S3
- Downgrading to DRBD 8.3
- Fencing in Libvirt/KVM virtualized cluster nodes
- Fencing in VMware virtualized Pacemaker nodes
- GFS2 in Pacemaker (Debian/Ubuntu)
- Interleaving in Pacemaker clones
- Maintenance in active Pacemaker clusters
- Managing cron jobs with Pacemaker
- Mandatory and advisory ordering in Pacemaker
- Migrating virtual machines from block-based storage to RADOS/Ceph
- Network connectivity check in Pacemaker
- OCFS2 in Pacemaker (Debian/Ubuntu)
- Solid-state drives and Ceph OSD journals
- Solve a DRBD split-brain in 4 steps
- Testing Pacemaker clusters
- Totem "Retransmit List" in Corosync
- Turning Ceph RBD Images into SAN Storage Devices
- Which OSD stores a specific RADOS object?
- Presentations
- Die eigene Cloud mit OpenStack Essex (German, LinuxTag 2012)
- Fencing (LCE 2011)
- GlusterFS in HA Clusters (LCEU 2012)
- GlusterFS und Ceph (German, CeBIT 2012)
- Hands-On With Ceph (LCEU 2012)
- High Availability Update (OpenStack Summit Fall 2012)
- High Availability in OpenStack (CloudOpen 2012)
- High Availability in OpenStack (OpenStack Conference Spring 2012)
- Highly Available Cloud: Pacemaker integration with OpenStack (OSCON 2012)
- Mit OpenStack zur eigenen Cloud (German, CLT 2012)
- Mit OpenStack zur eigenen Cloud (German, OSDC 2012)
- More Reliable, More Resilient, More Redundant (OpenStack Summit April 2013)
- MySQL HA Deep Dive (MySQL Conference 2012)
- MySQL High Availability Deep Dive (PLUK 2012)
- MySQL High Availability Sprint (PLUK 2011)
- OpenStack Essex im Praxistest (German, Linuxwochen Wien 2012)
- OpenStack High Availability Update (Grizzly and Havana)
- Roll Your Own Cloud (LCA 2011)
- Storage Replication in HPHA (LCA 2012)
- Zen of Pacemaker (LCA 2012)
- Technical documentation
- News releases
- Hints and Kinks
- What we charge
- What others say

"DRBD" is a registered trademark of LINBIT Information Technologies GmbH.
Downgrading to DRBD 8.3
A popular question to pop up on the drbd-user mailing list is how to downgrade from a DRBD 8.4.x installation to DRBD 8.3. Luckily, the developers haven't made this hard; there's just one specific thing you'll have to remember. We'll show you how to do this.
The DRBD metadata format between is largely compatible between DRBD 8.3 and 8.4, however there has been a change to the Activity Log (AL) – and that means that DRBD 8.3 can't immediately use a metadata set that was previously created under, or converted to, DRBD 8.4.
Thus, if your AL is cleanly "applied", then DRBD 8.3 will load the metadata just fine – if it's not, then it'll fail to attach.
There are two ways of fixing this. Note that both of them require that you bring the entire DRBD resource (both peers) down temporarily; there is no provision for a rolling downgrade. So be sure to schedule a maintenance window if this affects a production system.
The apply-al trick
This method is for you if you still have DRBD 8.4 utilities and the kernel module installed, before you start the downgrade procedure. Here's how it works; all of the steps listed here have to be completed on both nodes:
-
drbdadm down all drbdadm apply-al all modprobe -r drbd
-
Uninstall DRBD 8.4 utilities and kernel module.
-
Install DRBD 8.3 utilities and kernel module.
-
Fix your DRBD configuration to clean any DRBD 8.4 specific syntax. In essence, apply the considerations from the Recent Changes to 8.4 in reverse.
-
Check your configuration with
drbdadm dump all. -
modprobe drbd
- Examine
/proc/drbdto ensure your kernel has in fact loaded 8.3. -
drbdadm up all
No other changes should be necessary to your configuration.
The create-md/new-current-uuid trick
This is the backup option; note the the apply-al method remains preferred.
This method is for you if you've already uninstalled your DRBD 8.4 packages, you can't be convinced to reinstall them, and you're certain that your peers were in sync when you downed them. The last part is really important, don't do this if there is even the slightest chance that they were out of sync.
In essence, this amounts to freshly reinitializing the DRBD metadata, and then skipping the initial sync as you already know they peers don't need to be synchronized. Here's how it works:
-
Install DRBD 8.3 utilities and kernel module.
-
Fix your DRBD configuration to clean any DRBD 8.4 specific syntax. In essence, apply the considerations from the Recent Changes to 8.4 in reverse.
-
Check your configuration with
drbdadm dump all. -
drbdadm create-md all
-
modprobe drbd
- Examine
/proc/drbdto ensure your kernel has in fact loaded 8.3. -
drbdadm up all
-
Your devices will now come up as
Secondary/SecondaryandInconsistent/Inconsistent, as any freshly initialized DRBD device. -
On only one node (really, don't do this on both) issue this command:
drbdadm -- --clear-bitmap new-current-uuid all
- Your devices should now still be
Secondary/Secondary, but their disk states should have changed toUpToDate/UpToDate.
Now, you're able to make your devices Primary again, and put them back under Pacemaker management.
- Anmelden oder Registrieren um Kommentare zu schreiben
