HP Elitebook 830 Gen 6: Issues Capturing Images and Deploying Images
-
@rocksteve69 Try
nvme_core.default_ps_max_latency_us=0
For the purposes of FOS, we don’t require a threshold where the drive assumes a lower power state, so we may as well try and eliminate lower power states altogether.
There may also be firmware updates for the SSDs in question
-
@Quazz said in HP Elitebook 830 Gen 6: Issues Capturing Images and Deploying Images:
nvme_core.default_ps_max_latency_us=0
Hi, I have tested with the command nvme_core.default_ps_max_latency_us=0 and it made no difference to the G6 with the Samsung SSDs
We see about the firmware update for the SSD and report back
-
@rocksteve69 Can I ask which model it is specifically?
-
@Quazz The SSD model is Samsung MZVLB256HAHQ-000H1-S425NE0M726296.
-
@rocksteve69 This may be too much to ask, but if you swapped the WD drive for the Samsung drive in the same model laptop, does the problem move with the drive? That would then point directly to the Samsung drive.
@Quazz do you think the updated version of partclone would solve this issue ?
-
@george1421 To be honest, I never expected the partclone update to really change anything in that regard. I just figured there was a possibility that it could improve somewhat.
I’ve been confusing this thread with another one I think. I assumed he was already on those inits.
It’s worth a shot!
Can be downloaded from: https://dev.fogproject.org/blue/organizations/jenkins/fos/detail/master/95/artifacts
-
I have done just that. If I swap the drives works fine with WD. Samsung doesn’t work within either laptop. I could attempt to put one of the drives within a known working G5 as this will completely rule out anything else.
-
I can try. How does one install the update version of partclone?
-
I have also just tried the Samsung G6 SSD in a G5 and it worked fine with no issues, imaged in record time.
Starting to get very confused.
-
@rocksteve69 The updated version of partclone should be in the link I provided below. Partclone is in the init.xz file. You would just download the init.xz and rename it and and move it to /var/www/html/fog/service/ipxe directory. Then for that specific host go into the host definition and under the init field.
Byt @Quazz looking at the link you provided earlier, I though you named this beta init something else than init.xz?? Is that link still correct?
-
@george1421 This one is auto-built by Jenkins using the github repository, the other one was the one I made for testing, but wasn’t fully complete (missing progress update patch)
Other than that they’re identical
Once you have the init.xz you can rename it to init_partclone.xz or something and then copy it to
/var/www/fog/service/ipxe
Then you can change the host init in the WebGUI at host settings to init_partclone.xz or whatever you named it.
-
@Quazz Apologies for the late reply, been away out the office.
We have confirmed we are using the latest version of the Kernel.
Using the developer version of FOG
Using the version of PartClone detailed in the link below.All of the above has no effect on the image deployment.
Now we have another spanner in the works since I have been away. We receieved another batch of G6’s now with a third manufacturer, Hynix, and the same is happening to them SSDs as well.
Currently speaking to HP as the Samsung SSDs in the G6 are an OEM disk and Samsung don’t list a firmware update for it and it seems neither does HP. So this maybe a dead end.
-
Also just to add. I have taken a G6 with a Samsung and a G6 with a Hynix, placed the drives in a G5, imaged them fine. I then placed the drives back into the original laptops and captured an image of both laptops with no issues at all at full speed.
-
@rocksteve69 This reminds me of an issue another user had on certain SSD/PC combinations where he had to create a working NTFS partition before tools could copy to it full speed. (not FOG specific)
Still don’t know how or why that’s a thing, and the creator of partclone seems to have given up in that direction too.
-
Yeah its just crazy and strange how if I then redeploy the image back to the same G6 it slows down to a halt and never finishes.
We have resorted to a temp work around of having to take the SSD out of the G6 and image them in a G5 and placing them back in. But having to built out 50 laptops over the next week is taking sometime lol.
-
@rocksteve69 I’d be interested in seeing if you’d see better results with a RAW capture.
-
Now even more confusing is that the image capture of Samsung in a G6 after been in a G5 is now working in other Samsung G6s! I have managed to image 4 laptops so far.
Not so lucky with the Hynix with the same method. Will attempt a RAW capture.
Also I have noticed that these new images within FOG are showing as 0 ib size. But there most be data as I have successfully deployed the Samsung image.
-
@rocksteve69 I’m wondering if clonezilla would give you the same results. In this idea use fog to image a G5 with the G6 image. Then turn right around without booting into the windows OS grab an image with Clonezilla and see if you can deploy a clonezilla captured image to a G6 with the impacted disk. Does it image at an acceptable rate? This would be a way around the issue until we can better understand the problem.
-
@rocksteve69 said in HP Elitebook 830 Gen 6: Issues Capturing Images and Deploying Images:
Also I have noticed that these new images within FOG are showing as 0 ib size.
FWIW: This in a way is only a label that is updated on a successful upload. The size is added during the capture process and is only used for display. It has no impact on imaging.
-
Downloading Clonezilla to give it a test.
Interesting though when I try to capture an image of a G6 with a Hynix it goes through the motion up until Partclone starts takes no data and states capture completed successfully and reboots within a few seconds. No errors reported and reports back that the task is completed to FOG which clears the task.