unevent

Members
  • Posts

    903
  • Joined

  • Last visited

Posts posted by unevent

  1. Exploring utilizing a ZFS pool for a Windows Server VM for data.  I've seen in the docs how to assign a VM to a physical disk, but is it possible to assign a ZFS pool to a VM vs a virtiual disk (single file)?  Say I want the OS to be a RAW virtual disk not on the array, but want the Server VM to have storage to it's own ZFS vdev pool vs. just a single file virtual disk.  Possible?  If so, can someone provide the details?  With multiple ZFS pools being a thing know, being able to have a ZFS pool to each VM would be useful.  Thanks

  2. Please move post as needed.  I have a few questions regarding the phone home/cloud ability of 6.10.x releases. Is the phone home function on by default or you have to install the My Server plugin? Do I need to create unraid.net credentials if I would like to keep my server offline and not be apart of the cloud capabilities? Another way of asking, are the new credential and online/cloud functions and other requirements off by default and need to be turned on if you want to use the features? Thanks

  3. 23 hours ago, Squid said:

    You can create a feature request if you like as @dlandon suggested, but it is unobtrusive (but noticeable), dismissible, and doesn't generate any other notifications unless you're in the page (and UD shares real-estate with Main)

     

    Not sure if this falls under a feature request though  There is already a system in place on unRAID to handle plugin updates on a schedule or manually.  Seems the issue is more the plugin authors or perhaps the unRAID team are not enforcing plugin protocols.

  4. 21 minutes ago, Squid said:

    The whole point behind the update banner (which you can easily dismiss for UD if you like) is to let the user know that there is an update available to the plugin whenever you are on the plugin's page.  This is handled outside of the "check for updates" settings, and no notifications are ever sent out by the banner appearing.

     

    This is because as a general rule, the only versions of any random plugin that are actually supported is the latest version.  Otherwise, the first thing on any support question for any plugin the author is immediately going to state "Update the plugin".

     

    If you do decide to dismiss the banner, then the next update to the plugin will trigger another banner to appear.  This is by design.

     

    Problem is it triggers when I access the GUI, which is the main page.  Unraid core does not do this when there is an update unless you click the appropriate controls to do a manual or automatic check on a schedule.  I don't see why plugins cannot be the same.  The notification settings (under settings) have plugin update check turned off so why not have all plugins honor that setting?

  5. On 4/19/2022 at 8:48 AM, Espressomatic said:

    Can anyone recommend a definitive way to hide update notifications for Unassigned Devices along the top banner in the browser?

     

    In Notification settings I've already un-checked all Plugin notifications but this doesn't seem to have had an effect for the notifications from Unassigned Devices.  This plugin is set to auto-update, but more often than not, I fire up the Unraid web UI between updates and  see these banners regularly.

     

    To bump the post, is there a way to suppress plugin update banners?  Thanka

  6. On 4/8/2020 at 3:47 PM, whipdancer said:

    I use paid zoho - $36 for 3 primary accounts, each with multiple aliases.  So my home domain, and my family domain are both available to send and receive, my dev related domain, and my 2 app related domains are on a 2nd paid account, all received in the zoho email app on IOS or web client.  Using the zoho email client on my iphone, I can also send from any of my email addresses.  Using the default IOS mail client (really, anything other than the zoho mail app), I can only send from the "main" email address - so far, although I'm still digging into this limitation.

    Mxroute looks interesting, but it doesn't actually look like it will save me any money.

    I looked at Gsuites, but we don't need the rest of the features because we get Office 365 for free (for now).

    Is that per year?  Was not aware Zoho did email aliases, they only list domain aliases in their pricing.  Seems odd since they charge per user, assumed they would want to limit alias capability.  Happy with mxroute so far, got on a lifetime deal for $99.  Unlimited domains and email I assume for as long as they are in business or don't change the plan.  They seem to be small and can be a bit aggressive in locking down abusers so their website wording reflects that.  Other than that, I have a couple domains there with about ten email accounts so far.  Domain DNS configuration would have been a pain as they don't really offer much in the way of instructions, but after setting up with Zoho on a trail basis got it all figured out using their instructions for DKIM, SPF, etc. config.  Edit: Without doing the lifetime promo, their lowest tier is $40/year for 10GB storage and unlimited domains and unlimited email accounts.

  7. To update my search results for email hosting, I have signed on with mxroute for my email needs.  They offer unlimited domains and unlimited email users with their plans.  Pretty good experience so far, have two domains set up with email.  Thanks again to those that replied.

  8. Just an FYI for those running BOINC with Snort/Suricata, apparently couple days ago a new rule was published that was blocking COVID-19 transfers with BOINC

    ET INFO Suspicious GET Request with Possible COVID-19 URI M1 

    I hadn't had any work on two machines in two days and the transfer queue was in continuous retry.  Seems there is some malware going around which is why that rule was created and propagated.  This particular case was directly tied to BOINC as the trigger so I have suppressed it and now back to crunching.  If you are out of work units and the transfer queue is full, something to look into depending on what monitoring you have on your network.

    • Like 1
  9. Thanks all.  GSuites a bit much for home use ($6/user if I read it right).  Going with Zoho for the short term even though it is more limited than I had with godaddy hosting.  Free version has 5 users, but with only one having pop/imap access, the others are webmail only.  Upgrade is $1/user to give all pop/imap access paid yearly so $60/year for 5 users which is ridiculous.  Will look into vps and see about rolling my own and if its possible.

  10. 6 minutes ago, whipdancer said:

    I use zoho. G-Suites from Google was an alternative I looked into.  None of the others I looked at were close in price.

    For Zoho, you using the free or paid?  If free, do you have access to pop/imap or just webmail?

  11. Place I am with now is getting expensive ($11/month) to host a domain for just email.  Looking for suggestions for hosts that allow to bring in my own domain and have ability to host email with multiple non-business users/addresses.  Thanks

  12. Drive issue not related to 6.8.2 upgrade as I did xfs check on all drives in maintenance mode before bringing array up after upgrade.  Got a read error on disk 2 on array start so I rebooted.  Drive came up and normal, but of course it is now disabled.  No smart errors so just something the repair did (but no lost+found either on that disk either).

     

    Only options I have in 6.8.2 to bring the array online is to do a 'read check' which is a no-correct parity operation.  Question is why does it force this as the only option instead of allowing a disk rebuild on the disabled drive?  This seems new as I don't recall seeing this behavior on 6.7.2 and older.  I need to rebuild the disk not read the emulated disk contents, or am I missing something?  Thanks

  13. Any idea what this might be?  I was not up at 2am using ssh and no ports are open to unraid gui.  Except for those two entries the logs are clean going back to the 15th.  Doesn't look like a hack since only two attempts of whatever this is in almost nine days of uptime.  Is this an internal ssh thing or something external triggered these log entries?  "kex_exchange..." didn't net much in a search.  Thanks

     

    Jan 23 02:00:31 Tower sshd[13206]: error: kex_exchange_identification: Connection closed by remote host
    Jan 23 02:00:32 Tower sshd[13281]: error: kex_exchange_identification: Connection closed by remote host

     

  14. Not sure if reported yet.  When formatting ext4 says 64-bit filesystem support is not enabled:

    Jan 13 20:32:06 Tower unassigned.devices: Format disk '/dev/nvme0n1' with 'ext4' filesystem result: mke2fs 1.45.0 (6-Mar-2019) 64-bit filesystem support is not enabled.  The larger fields afforded by this feature enable full-strength checksumming.  Pass -O 64bit to rectify. Discarding device blocks:      4096/250050816#010#010#010#010#010#010#010#010#010#010#010#010#010#010#010#010#010#010#010  629555{.....}done 

     

  15. 14 hours ago, trurl said:

    When doing filesystem repair you should capture exactly your input and the resulting output so you can post them.

    I have the logs.  Just had no need to post drive contents to the world.  Lost 1.25TB on this drive.  Perhaps a feature request is to halt parity when xfs reports first error so instead of parity protecting the corruption it keeps the parity intact as much as possible.  Seven plus years with unRaid and first case of corruption and with xfs and data loss.  No power loss, no unclean shutdowns.

  16. Thanks for the help.  I had started array in maintenance mode and still showed unmountable.  Remember stories inthe past about starting array with a troubled disk and loosing data.  Started array normally and now sorting through lost+found to see what I need to replace.  Thanks again.

  17. Ran xfs_repair without options and completed on disk2.  Will look in lost+found later today to see what was trashed.  What would be recommended path to get back online at this point?  Never had corruption and been using since 4.7 days.  Never even had parity sync issue.