unevent

Agree it is the link being reset, but it occurs after the start of the default script - repeatably, but not with the Joe L script. Speeds started around 200MB+ then slowed to 10-15MB after 20-30 seconds and the link/drive issues started spamming the log. I stopped the preclear, powered off the external dock (JMicron JMS561 chipset), removed the hard drive, re-inserted, powered it back up and let it link up and stabilize. After a couple minutes idle I started the default script again and got another bunch of errors - attached as '...2nd'. Stopped the preclear then switched to Joe L script and started - no other changes and no touch to hardware. Worked fine as the attached shows and current status is below. I agree it looks like standard SATA cable or power issue, but this is an external eSATA dock with its own power supply and the problem persisted after drive removal/attach and cable reattach. When this preclear is done I plan to try the default script again with no other changes or touching of hardware. What else concerns me is the start sector of 1 vs 64 which is what the Joe L script should have defaulted to. Never seen one where it was set to 1 and don't think that was ever an option with the original script - was either 63 or 64. After this default script test I may install screen and do the original Joe L script like the old days and see what happens. ================================================================== 1.15 = unRAID server Pre-Clear disk /dev/sdq = cycle 1 of 1, partition start on sector 1 = Disk Pre-Clear-Read completed DONE = Step 1 of 10 - Copying zeros to first 2048k bytes DONE = Step 2 of 10 - Copying zeros to remainder of disk to clear it DONE = Step 3 of 10 - Disk is now cleared from MBR onward. DONE = Step 4 of 10 - Clearing MBR bytes for partition 2,3 & 4 DONE = Step 5 of 10 - Clearing MBR code area DONE = Step 6 of 10 - Setting MBR signature bytes DONE = Step 7 of 10 - Setting partition 1 to precleared state DONE = Step 8 of 10 - Notifying kernel we changed the partitioning DONE = Step 9 of 10 - Creating the /dev/disk/by* entries DONE = Step 10 of 10 - Verifying if the MBR is cleared. DONE = Post-Read in progress: 4% complete. ( 337,150,976,000 of 8,001,563,222,016 bytes read ) 203 MB/s Disk Temperature: 36C, Elapsed Time: 29:21:49

Seem to be having an issue with the default script vs the Joe L. script. I purchased a new Toshiba N300 8TB NAS drive and started preclear using Preclear plugin version 2017.09.27 with default gfjardim script, clear, no notifications, nothing skipped. After about 20-30 seconds I had errors spamming syslog. Stopped preclear ran smart on it and no errors. Removed drive from my eSATA drive dock and reinstalled and powered it up. No abnormal noises before or after. Started preclear again with same settings, same log spam. Stopped, switched to the Joe L. script (the modified for notifications version) and started with defaults. Seems to work fine, no errors in log after 20 minutes. Log excerpt and preclear status attached. ToshibaN300_JoeL_PreclearLog.txt ToshibaN300_Log-1st.txt ToshibaN300_Log-2nd.txt ToshibaN300_JoeL_PreclearLog_26m.txt

Thanks.

Is the Splunk Docker still being maintained/updated?

To throw another option out, look at the ASRock J3355B-ITX. It has an embedded Celeron J3355 which is dual core 2.0GHz, w/AES-NI and 10W TDP. In the US it goes for around $55.

Plex has removed the option to completely opt-out of their data collection and have changed what they collect as well. Their method has changed quite a bit since they first announced they were doing this. It was a simple act of blocking 'metrics.plex.tv' from the server, but have since continued to move to the client for data collection. They do have a limited opt-out still available on your plex account on the website. It is rather simple and confusing as the page you view just has a single tick-box in the middle with no submit or other buttons. Just un-ticking (or ticking) saves it. For those that don't like this way of doing business, recourse options appear to be nil, but common-sense suggests don't log in using social media credentials or link to social media and have all clients opt-out on the individual accounts. The harvested data is not in the clear so most likely, as with most businesses in the market of harvesting data, someone will hack it and find out they are collecting more than the policy wording says they are. Happens all the time. I hear Emby is becoming quite popular because of these changes with Plex.

Ooooooooooo Classic! 16-bit. Got DOS 6.22?

Can try the Virtualbox plugin for unRAID. Can't have Virtualbox and KVM enabled at same time, so one or the other.

@tjb_altf4 brings up a point, is your download/data folder cache-only? Only the final copy of the finished and processed download should be to the array.

You will need to be in the Docker container to execute this script. Copy the 'configurator.py' script to your appdata config directory for the home assistant Docker. Before moving forward, you need to use a Linux text editor to edit the script and configure for your environment. Using 'mc' Midnight Commander on unRAID via Telnet/SSH is one way. Telnet/SSH to unRAID and issue the command below replacing 'container_name' with the name of the container as seen on the Docker tab in unRAID GUI: docker exec -ti container_name /bin/bash Assuming your folder map for config is '/config', verify the existence of the configurator.py script in the config folder: ls /config/conf*.py After verifying/locating the file, make the script executable: chmod +x /config/configurator.py Execute the script: cd /config ./configurator.py Follow the instructions from the Hass.io site.

Reading up on this script it seems it provides a web-based interface to the configuration. Will need to know what Docker you are using to determine if you can use this script.

Telnet/SSH to unRAID and issue the command below replacing 'container_name' with the name of the container as seen on the Docker tab in unRAID GUI: docker exec -ti container_name /bin/bash Once in the container, issue these commands to verify nice and ionice are available: nice --version ionice --version

Depending on the Docker you are using, you can use 'nice' and 'ionice' settings to tame Sab with regards to CPU and disk priority. Sab wiki link. If the Docker has nice and ionice then these settings will work.

Let's rewind this a bit. Can you explain what you are trying to accomplish? This script, is it part of a Docker? If so which one? Explain what you are trying to accomplish and the forum will be able to better help you. What I have seen so far is you are trying to run a python script 'in the open' on unRAID. This is bad.

There is no half duplex for gigabit - only on 100 and 10Mbit. Should not even have the option to manually specify duplex for gig, should be set to auto and be the only choice. Post a diagnostics and also describe your network path and hardware to your Xfinity connection. Also explain what you are able to do on 'full blown Linux boxes'.

Only interest I have is the effect it will have on CPU prices. Intel coming out with brand new 8th gen at around $380 for their high end offering...unheard of before AMD got their groove back.

The StarTech 2 Port card uses Marvel 88SE9128 chipset. Forum search The StarTech 4 Port card uses Marvel 88SE9230 chipset. Forum search Typically Silicon Image and Asmedia chipsets have better support, Marvel are hit and miss.

That script is requiring python3 (first line, #!/usr/bin/python3). Not installed on unRAID by default and need to use the Nerd Pack to get it. This part of a plugin or Docker?

Yep Quantenna, thanks for the clarification.

+1

If you are saturating a 100Mbit WAN connection 'Captian Obvious' should have kicked in long before in the hardware selection process. For your Asus, they moved to Qualcomm vs Broadcomm for that model and the 5GHz was limp. Look into Merlin firmware for that model, there are some improvements over stock.

These appear to be older drives that have been phased out. They are seven platter OEM bare drives that came out around 2014 with a five year original warranty, yet being sold by a third party wholesaler/reseller through Newegg with a three year warranty. The vendor lists them for $195 on their own website with a 5 year warranty. goHardfDrive is not an authorized distributor for HGST and without a serial number not possible to verify a warranty with HGST. For $10 more you can get a HGST Deskstar NAS retail box (vs the Ultrastar He6 OEM) 6TB at B&H. Amazon, Fry's and Newegg also had the sale, but no longer.

Depends on what the operations are. Small stuff probably not noticeable.

During preclear you are either writing to the drive as fast as it can or you are reading from the drive as fast as it can (it being the drive). The 'dd' command being executed is basically always waiting for the drive. Someone else can probably explain better.

With NUT not installed and the USB cable to the UPS connected to the server and apcupsd enabled as usb, usb, do you get status of the UPS in the GUI? Is everything working, but you get these communication lost messages? This change in cable management, how is the USB cable to the UPS ran now versus what it was before? Is it bundled with other cables like power?