Low speed second replication VM

[Iven]

version 6.5.0.128, vmware esxi 5.1, vmx-09, virtual machine - 150 gb, 50% - free space, local replication from one hdd on other
core i7, ssd, 16gb

i created replication job
First clean replication job took 1 hout 30 min
I made second replication immediately after the first job - no any change on server - it took about 20-30 min
I made next replication after 8 hours of working (terminal server windows 2003, 10 users)
i could not finish it becouse this replication job took 10 hours for about 50-60% of replication
i saw speed of replication about 500kb-1mb
why so slow ?

[foggy]

Iven, what are the bottleneck stats for this job run, what transport mode is being used to retrieve VM data from the storage (available next to the proxy name selected for processing), is CBT enabled on this job, what are the data read/transferred counters?

[Iven]

all settings is default
yes, cbt is on
transport mode is being used to retrieve VM data from the storage (available next to the proxy name selected for processing)
hm cant find

processing rate for first replicating was about 13-20 mb
for all other about 512kb-1mb

Name: Replication Job 1
Type: VMware Replica
Source items:
SKLAD (192.168.30.239)
Replica destination:
Host: 192.168.30.239
Resource pool: Resources
VM folder: vm
Datastore: datastore2
Replica suffix: _replica
Restore points to keep: 3
Source proxies: Automatic selection
Target proxies: Automatic selection
Repository for replica metadata: Default Backup Repository (host: This server, path: c:\backup)

[foggy]

Please open job statistics window (right-click the job and select Statistics). What is the main bottleneck stated there? Please select the particular VM in the list to the left and look for the transport mode tag next to the proxy server name (should be [nbd], [hotadd], or [san]). Also, is the [CBT] tag displayed next to the VM disks being processed?

[Iven]

i see in statistics Summary, Data and Status
In Summary i see bottleneck Target
Source 2%
Proxy 28%
Network 1%
Target 99%
this information from first clean replication

i dont see any proxy server name
where i must to find it ? option of replication job ? which of settings ? name, virtual machine, destination, job settings, guest processing, schedure, summary ?
Or window Virtual Machines ? i see there only Vmware vPshere and one my esxi server without any usefull information in properties

[foggy]

Target 99%
this information from first clean replication

That means that the target disk writer component spends most of its time writing data and is not able to accept data any faster. Does proxy server has access to the target datastore? What repository is being specified in the replication job for storing replica metadata?

Proxy server name is available in the job session log if you select the particular VM in the list to the left. Job session log is available either by right-clicking the job and selecting Statistics (click Show Details for the full view) or by selecting the job and then selecting View > Full View.

[Iven]

are u sure that we are speaking about 6.5 ?
i havnt Proxy servers names in statistics
i havnt proxy server deployed on the target host - only vmware esxi 5.1 on ssd with vmx-09 without any raid or online repositore or vmware center or else
one virtual machine with veam, second with target server on one ssd datastore 1, and replication machine on second hdd datastore 2

What repository is being specified in the replication job for storing replica metadata?
as i said - all default
all Automatic selection
defaul beckup repository created byt veam beackup

[foggy]

Could you please post two screenshots of the job statistics window (general view and when the replicated VM is selected in the left panel)?

[Iven]

different statistics ? i cant find second

[foggy]

Please select the SKLAD VM in the window displayed on the first screenshot. You will see the VM processing log with the details required. I also would like to see the same screenshot for the incremental job run, which actually causes the issue (since the speed for the initial replication is quite acceptable).

[Iven]

oh, sor
now i understand what u wanted )

[foggy]

Iven, could you please try to perform another full to see whether this behavior is specific to incremental runs only? Could there be some infrastructure changes performed after the previous full replication that could negatively affect target storage?

[Iven]

no any infrastructure changes
it is done every night
btw this is the first time when it was over for the night
because the first work was done in the afternoon instead of the morning or yesterday
as u see it transferred 2.6 gb changes
i tried delete replication jobs and create new but first replication continues always less then 2 hours, but it is too long after a day of work
that's one of the not finished
I stopped it manually

[foggy]

At that point I think you'd better contact technical support directly as further troubleshooting via forums does not seem to be effective.