I would like to use this newsletter edition to focus on a few upcoming events and conferences. Events have always been important to LINBIT, as I’m sure they are for many companies and people across the open source community, and we have some news to share.
Coming up first on March 6-9, LINBIT will be at the SCaLE 22x event in Pasadena, California. SCaLE is the largest community-run open-source and free-software conference in North America, and we’re proud to be one of the conference sponsors. If you find yourself in attendance, we’d love to meet. You can find us at booth 319.
Next, and something we eagerly look forward to, LINBIT will sponsor and host the CloudStack European User Group 2025 community conference in our home city of Vienna on May 8. This announcement is quite timely, as it comes shortly after we released the LINBIT CloudStack HCI Appliance, allowing users looking for a VMware alternative to set up CloudStack in under 20 minutes.
The LINBIT team has been actively developing for the CloudStack platform since early 2021. However, the LINBIT CloudStack HCI Appliance takes things to a whole new level by removing all the complexity of installing and getting started with CloudStack. We’re looking forward to the event, which, as always, is an excellent opportunity for the community to come together.
May promises to keep us busy with one additional event scheduled before the month is through. We are exhibiting at the Red Hat Summit 2025 in Boston, Massachusetts, May 19-22.
Over to internal software updates, on the DRBD side, we are working on a new, alternative way of doing a resync when the backing block device of the resync-target / Secondary side limits performance. That might be the case when the Secondary has inferior hardware or when you run a central secondary for many primary nodes.
The current resync implementation does the resync in parallel with online replicating application writes. The idea for the new, alternative implementation is to do the resync before enabling online replication. With that, we need to do multiple iterations of the resync since application writes might bring the replicas out of sync again.
After each pass, the algorithm checks if the out-of-sync number decreases and decides when to enable parallel online replication. This method might positively affect many scenarios, as it can shorten resync time. I look forward to seeing the first results from real workloads soon.
Regarding the latest software updates, the team has released linstor-server 1.30.4 and drbd-reactor v1.8.0.
