Cold migration profile: OpenVDDK and VMware VDDK 8.0.3
These runs were recorded on 4 October 2026. At that time every name lookup on our lab's workers waited 5 s for a DNS server that did not answer. VMware's library made four lookups in its disk open and one in each of its seven allocated-block queries; OpenVDDK's disk open and queries made none. V2K Migrate's own vCenter connections waited the same way in both runs. Closing the disk, reading and the copy phase were not affected.
Plan durations: OpenVDDK 444.845 s; VMware VDDK 906.308 s. Both include setup and waiting; VMware's includes the name-lookup waits described above.
Copy phase of the 200 GiB disk, from the application's progress lines: OpenVDDK 381 s (537 MiB/s); VMware VDDK 727 s (282 MiB/s).
One Windows Server 2019 VM with one thin-provisioned 200 GiB disk: 181.3 GB of data copied and 33.4 GB of zeros skipped. Cold migration from a NetApp NFS v3 datastore to an NFS export on the same NetApp, with one worker VM on the source VM's ESXi host. A later run from an NFS 4.1 datastore is not included.
Each run is shown from the moment its migration plan started until it completed, and both graphs in a row use the same time axis. The grey area in the OpenVDDK graphs is the time after that run had already finished. Averages and 95th percentiles cover the whole plan, setup and waiting included; read latency counts only the moments when the disk was being read. They are not copy-phase metrics; VMware's include the name-lookup waits, which lower its averages. Runs on 4 October 2026: VMware VDDK 01:55–02:10, OpenVDDK 02:52–03:00 (CEST).
Both runs were sampled every two seconds inside the worker VM and on the ESXi host; the moment the source disk appeared in the worker and started being read was sampled every 100 ms.
OpenVDDK
VMware VDDK 8.0.3
Migration plan, start to finish
444.845 seconds (7.41 minutes)
02:52:57–03:00:22 CEST
Migration plan, start to finish
906.308 seconds (15.11 minutes)
01:55:10–02:10:16 CEST
From disk attached to the first 8 MiB read
Pass
Disk visible in the worker
Time to the first 8 MiB read
Cold copy
02:53:39
0.402 s
Sampled every 100 ms. Small reads while the disk is identified are not counted. This is the read progress seen in the worker, not the library's Open call.
From disk attached to the first 8 MiB read
Pass
Disk visible in the worker
Time to the first 8 MiB read
Cold copy
01:56:30
52.753 s
Sampled every 100 ms. Small reads while the disk is identified are not counted. This is the read progress seen in the worker, not the library's Open call.
This interval includes VMware's seven allocated-block queries, each with one 5 s name-lookup wait.
Inside the worker VM
Storage network, sent
Mean 390.09 · p95 575.27 · Peak 635.50 MiB/s
Storage network, sent
Mean 191.36 · p95 346.24 · Peak 409.80 MiB/s
Storage network, received
Mean 1.73 · p95 1.53 · Peak 108.20 MiB/s
Storage network, received
Mean 0.85 · p95 0.91 · Peak 117.89 MiB/s
Worker CPU busy
Mean 37.83 · p95 49.49 · Peak 54.97 %
Worker CPU busy
Mean 18.39 · p95 26.64 · Peak 33.22 %
Worker CPU waiting for I/O
Mean 39.99 · p95 59.60 · Peak 63.74 %
Worker CPU waiting for I/O
Mean 18.21 · p95 25.73 · Peak 27.27 %
Worker CPU steal
Mean 0.00 · p95 0.00 · Peak 0.00 %
Worker CPU steal
Mean 0.00 · p95 0.00 · Peak 0.00 %
Worker memory used
Mean 0.73 · p95 0.79 · Peak 0.86 GiB
Worker memory used
Mean 0.70 · p95 0.74 · Peak 0.78 GiB
Worker swap used
Mean 0.01 · p95 0.01 · Peak 0.01 GiB
Worker swap used
Mean 0.01 · p95 0.01 · Peak 0.01 GiB
Hot-added disks (system disk not counted)
Mean 0.87 · p95 1.00 · Peak 1.00 disks
Hot-added disks (system disk not counted)
Mean 0.89 · p95 1.00 · Peak 1.00 disks
Read rate of the hot-added disk
Mean 461.40 · p95 936.15 · Peak 966.90 MiB/s
Read rate of the hot-added disk
Mean 227.12 · p95 581.49 · Peak 616.13 MiB/s
Hot-added disk busy
Mean 82.69 · p95 99.25 · Peak 100.00 %
Hot-added disk busy
Mean 64.33 · p95 88.85 · Peak 90.86 %
Management network, received
Mean 0.01 · p95 0.01 · Peak 0.03 MiB/s
Management network, received
Mean 0.01 · p95 0.01 · Peak 0.19 MiB/s
Management network, sent
Mean 0.03 · p95 0.03 · Peak 0.04 MiB/s
Management network, sent
Mean 0.03 · p95 0.03 · Peak 0.04 MiB/s
Destination NFS mount
NFS 4.2, 1 MiB read and write size, hard mount, TCP
Destination NFS mount
NFS 4.2, 1 MiB read and write size, hard mount, TCP
On the ESXi host
Worker VM CPU ready (ESXi)
Mean 1.24 · p95 1.81 · Peak 2.40 %
Worker VM CPU ready (ESXi)
Mean 0.98 · p95 1.58 · Peak 2.38 %
Worker VM CPU co-stop (ESXi)
Mean 0.00 · p95 0.00 · Peak 0.00 %
Worker VM CPU co-stop (ESXi)
Mean 0.00 · p95 0.00 · Peak 0.00 %
ESXi host CPU used
Mean 12.85 · p95 16.87 · Peak 19.89 %
ESXi host CPU used
Mean 11.48 · p95 16.04 · Peak 21.64 %
Read latency per request at ESXi
Measured only while the disk was being read; very short reads can be missed.
Mean 18.79 · p95 23.52 · Peak 29.54 ms · 192 samples
Read latency per request at ESXi
Measured only while the disk was being read; very short reads can be missed.
Mean 2.51 · p95 3.81 · Peak 8.92 ms · 367 samples
ESXi storage network, received
Mean 443.10 · p95 871.62 · Peak 948.92 MiB/s
ESXi storage network, received
Mean 218.73 · p95 553.15 · Peak 599.95 MiB/s
ESXi storage network, sent
Mean 4.67 · p95 7.52 · Peak 13.57 MiB/s
ESXi storage network, sent
Mean 4.00 · p95 8.50 · Peak 39.28 MiB/s
Sampling
223 samples in the worker and 221 on the ESXi host during the plan
Dropped incoming frames: 58 on each network card; no errors or dropped outgoing frames.
453 samples in the worker and 451 on the ESXi host during the plan
Dropped incoming frames: 120 on each network card; no errors or dropped outgoing frames.