Poolsman - ZFS Web GUI for Linux based on Cockpit

Notice: Page may contain affiliate links for which we may earn a small commission through services like Amazon Affiliates or Skimlinks.

cojones

New Member
Jan 7, 2023
6
4
3
just checking in as I haven't seen any updates on the site, any news on a new release?
 

PoolsmanTeam

New Member
Sep 12, 2022
16
12
3
poolsman.com
just checking in as I haven't seen any updates on the site, any news on a new release?
We've released interim 0.3.5.0 version some time ago, that is available for download. It hasn't been publicly announced because included various small improvements and bug fixes, and has been addressed to users, who faced these issues (support of vdev_id.conf file, display DRAID pools in UI, better support for NVMe S.M.A.R.T., support for pools with listsnapshots=on property, etc). We are still working on new Preview 4 (0.4.0.0) release with the new set of features and currently expect it on March 6.
 
  • Like
Reactions: gb00s

PoolsmanTeam

New Member
Sep 12, 2022
16
12
3
poolsman.com
We are pleased to announce the 4th Poolsman preview release with significantly improved Pool Creation Wizard:
  1. Now you can add any number of disk groups (VDEVs, virtual devices) during pool creation. All types of disk groups (except very specific brand-new draid) are supported, including mirror, raidz, spare, log, dedup, special and cache.
  2. Ability to set specific pool options during pool creation with the possibility of furter editing. They include Sector Size, AutoTrim, AutoReplace and AutoExpand options (AutoReplace is available in read-only mode for now due to high risk of data loss in case of incorrect usage).
  3. Improved Pool Review page in Pool Creation Wizard, which includes high-level information about the new pool with the ability to preview the Shell Command, that will be executed.
  4. Added support for viewing topology of existing DRAID pools in read-only mode.
  5. New Pool Maintenace page, that currently includes pool export and destroy options (scrub, trim, upgrade pool options are on the road).
  6. Better S.M.A.R.T. support (based on users experience).
  7. Various small improvements and bug fixes (support for pools, that were configured manually with using of vdev_id.conf file, have listsnapshots option enabled, etc.)
 

Attachments

  • Like
Reactions: gb00s

PoolsmanTeam

New Member
Sep 12, 2022
16
12
3
poolsman.com
Regarding SMART and disks, how do you currently determine that a disk is healthy? Are you perhaps monitoring these? What SMART Hard Disk Errors Actually Tell Us
We use smartmontools to get the SMART data (the same package as in the article you mentioned). Poolsman mostly displays the data that it gets from this tool (the overall health state is also provided by this). Additionally we perform some minor transformations to make the data more friendly (e.g. calculate host reads/writes in bytes). You also can manually check any of the SMART attributes (including attributes from the article), provided by your drive and supported by smartmontools package.
Additionally, if you enable logging, you should be able to review shell commands, that Poolsman executes to get the SMART data (and any other data). You can see how it looks on the screenshot.
 

Attachments

cojones

New Member
Jan 7, 2023
6
4
3
Thank you. The reason I'm asking is because those specific SMART params seem to signal an impending drive failure without necessarily triggering SMART overall health alarms - backblaze have a massive statistic body of evidence on it. I think there's an opportunity here for SMART(er) drive health monitoring beyond simply echoing SMART values back to the user. Quoting the article:

1678700306639.png
 
  • Like
Reactions: gb00s

PoolsmanTeam

New Member
Sep 12, 2022
16
12
3
poolsman.com
Thank you. The reason I'm asking is because those specific SMART params seem to signal an impending drive failure without necessarily triggering SMART overall health alarms - backblaze have a massive statistic body of evidence on it. I think there's an opportunity here for SMART(er) drive health monitoring beyond simply echoing SMART values back to the user. Quoting the article:

View attachment 27864
Thanks. It definitely makes sense to monitor these particular SMART attributes in parallel (and/or highlight them when they RAW values are greater than zero). We've added a research task for that into our backlog. Probably we will add some kind of such monitoring in one of our next releases (at least simple highlighting of these attributes should be easy to implement).