Unable to start or reinstall node software after Windows 11 update

Node Operators troubleshooting
,
1

Hey all,

I have a Storj node that has been running normally for over a year on a small Windows 11 23H2 mini PC. I was prompted for a Windows update recently and went ahead and installed and rebooted the PC. Since installing the update (PC still says it’s on 23H2 post-update), I have been unable to start my storage node back up. Trying to visit the dashboard at localhost in my browser showed an error.

I got into the terminal and ran “.\storagenode.exe info” and received warnings that the software could not find several folders related to node operations like “temp,” and “blobs,” and that it was looking for them within C:\Users(Me)\AppData\Roaming\Storj. I had those specific folders set up on my hard drive for Storj network data, so it seemed that the mapping for my folders had been blown out.

I could not find the mapping for these folders within config.yaml, so I decided to uninstall my existing Storage Node installation and reinstall it. I did not delete my Storage Node install directory, the existing network data on my hard drive, or my identity files within AppData.

When I downloaded the latest version of the Storage Node GUI installer (modified date of May 5, 2026) I ran the installation process and chose the same directories as I had in my previous installation for install folder, identity and network data. Everything that I can visibly see in install progress runs fine up until it says that the installer is creating a new shortcut, and a terminal window briefly pops up before the installer displays this message:

“There is a problem with this Windows installer package. A program run as part of the setup did not finish as expected. Contact your support personnel or package vendor.”

I noticed that I have an 86-gigabyte log file, so I also tried moving my existing Storage Node folder to a separate backup folder and ran the installer again, but pointed the install directory to a new folder. The exact same error pops up during install progress and I am unable to complete it.

Any idea what I could try next to get back up and running? I am hesitant to check the log file because I don’t think my mini PC could handle processing it.

2

Consider switching to linux to solve this once and for all, instead wasting your time with toy operating system (Windows 11).

3

I do also have a Debian server, but my drives are formatted to NTFS and in a Windows-only software JBOD configuration. I also have over 6 terabytes of client data on them and switching would mean all of that gets deleted. I would like to preserve my existing node and keep this one going if I can.

4

does you paths names to folders contains spaces? it dont like it
i hold all identities on C:\identitie, C:\identitie1 …
and paths to data that also dont contain spaces

5

Not a Windows user myself, but it’s fine to move this file elsewhere (btw turn NTFS compression on it to save space), and just post whatever new entries will show up in a fresh file.

6

The problem is that no new logs are being generated, because the installer for Storj’s software is failing now. Unless the installer generates a log somewhere to figure out why it’s halting?

7

If so, you’ll probably have to wait for the staff to answer, sorry!

8

This is the default folder. Use param --config-dir to specify your folder.

9

I am unable to do so now because the Storage Node software is now failing to reinstall.

10

Did you run the installer as administrator?

11

The option to do so is grayed out under Compatibility in Properties, not sure why. Maybe because the installer is an .msi instead of .exe? I do get a prompt to continue as an administrator before the installation begins (once I’ve filled out all the details like network data directory). I did click yes to that and the installation fails shortly after.

12

No reason to keep guessing. Provide:

  • output of sc qc storagenode
  • output of sc query storagenode
  • exact full path to your existing config.yaml
  • the storage.path, identity.cert-path, and identity.key-path lines from that config
  • Windows Event Viewer entries from the failed installer attempt
  • MSI verbose install log

Run this from an elevated terminal (Log, *All, Verbose):

msiexec /i path\to\storagenode.msi /L*V %TEMP%\storagenode-install.log

Then post the section around Return value 3.

The node looking in C:\Users\<you>\AppData\Roaming\Storj means it is running with default config discovery, or the service/install arguments pointing at your real config are gone.

It makes absolutely no sense to cargo-cult it via “reinstall”, “run as admin”, reinstall windows, and other baseless voodoo.

For the record, I agree with folks above. Ditch this garbage parody to an operating system, reformat to ZFS and restore data from backup. The sooner you do that the less suffering will you endure overall.

Edit.

Actually, maybe ditch the installer failure investigation and create a new service manually.

sc create storagenode binPath= "\"C:\Program Files\Storj\Storage Node\storagenode.exe\" run --config-dir \"D:\whatever\storj\"" start= auto

I don’t remember if any additional parameters are needed.

Edit2: I haven’t used windows in over a decade, so I may be way off on syntax, which is adapted from my old notes.

13

Hello @MarkFeuer,
Welcome back!

why did you reinstall it in a first place? Reinstallation usually never helps.
To make an info command works you need to provide a path at least to the config file, as noted here:

The usual reason for failing the installer is not removed "C:\Program Files\Storj" folder and also the file

After removing them you should be able to reinstall, please specify the correct path to the identity.

14

Hey again @Alexey , hope you’ve been well!

Reinstalling was not the smartest move on my part, but I know for next time not to do that again. I thought reinstalling might help to map the directories for storage/identity/etc. again.

Deleting the “.trash-uses-day-dirs-indicator” file was the fix for reinstalling, once that was eliminated from my data directory the reinstall process finished without any issues! Thank you for your help with that.

Checking under Windows Services, I confirmed that Storj V3 Storage Node and the Updater are both running.

When I looked at the new log file after reinstalling, I saw this entry:

2026-05-17T13:53:27-05:00	INFO	hashstore	loaded hint file	{"satellite": "12EayRS2V1kEsWESU9QMRseFhdxYxKicsiFmxrsLZHeLUtdps3S", "store": "s0", "path": "I:\\Storj\\hashstore\\12EayRS2V1kEsWESU9QMRseFhdxYxKicsiFmxrsLZHeLUtdps3S\\s0\\meta\\hint-000000000000012e", "writable": [6260, 202, 7384, 5135, 6378, 7034, 7355, 7304, 7152, 7174, 600, 6868, 7237, 5034, 7222, 7301, 7040, 1741, 6883, 7161, 1064, 6954, 5134, 4877, 7112, 5657, 6685, 149, 7182, 661, 7385, 7011, 901, 2670, 5955, 6760, 6360, 6713, 6979, 5138, 7117, 4982, 6221, 7394, 7221, 6884, 6897, 5878, 7164, 7295, 1637, 7395, 6881, 7224, 5978, 2622, 7246, 6811, 6223, 131, 7278, 3181, 7025, 6905, 7035, 7315, 7121, 6460, 6245, 1678, 1265, 6294], "largest": 7395, "skip": false}
2026-05-17T13:54:15-05:00	INFO	hashstore	unclean shutdown detected: reconciling logs	{"satellite": "12EayRS2V1kEsWESU9QMRseFhdxYxKicsiFmxrsLZHeLUtdps3S", "store": "s0"}

A few minutes later, entries like these started popping up in the log:

2026-05-17T13:59:00-05:00	WARN	hashstore	mismatched log tail	{"satellite": "12EayRS2V1kEsWESU9QMRseFhdxYxKicsiFmxrsLZHeLUtdps3S", "store": "s0", "id": 7394, "path": "I:\\Storj\\hashstore\\12EayRS2V1kEsWESU9QMRseFhdxYxKicsiFmxrsLZHeLUtdps3S\\s0\\e2\\log-0000000000001ce2-00005069", "table": null, "log": [{"Key":"[REDACTED]","Offset":524866688,"Log":7394,"Length":54016,"Created":20570,"Expires":41171},{"Key":"[REDACTED]","Offset":524860224,"Log":7394,"Length":6400,"Created":20570,"Expires":41171},{"Key":"[REDACTED]","Offset":522540288,"Log":7394,"Length":2319872,"Created":20570,"Expires":41171},{"Key":"[REDACTED]","Offset":522536384,"Log":7394,"Length":3840,"Created":20570,"Expires":41171},{"Key":"[REDACTED]","Offset":522534016,"Log":7394,"Length":2304,"Created":20570,"Expires":41171}]}

Then this just popped up in the log as well:

2026-05-17T14:02:54-05:00	INFO	hashstore	loaded hint file	{"satellite": "12EayRS2V1kEsWESU9QMRseFhdxYxKicsiFmxrsLZHeLUtdps3S", "store": "s1", "path": "I:\\Storj\\hashstore\\12EayRS2V1kEsWESU9QMRseFhdxYxKicsiFmxrsLZHeLUtdps3S\\s1\\meta\\hint-0000000000000117", "writable": [7684, 8096, 7898, 7750, 7917, 471, 3861, 7970, 447, 7421, 8091, 7754, 6173, 7824, 7327, 290, 3539, 2549, 8169, 5533, 8015, 8111, 5479, 7683, 7919, 8168, 8171, 1752, 8122, 5728, 7899, 2964, 8110, 8098, 7331, 1237, 7966, 8017, 7876, 1065, 3051, 5679, 7903, 7878, 5241, 8174, 7897, 7428, 7301, 7880, 195, 7055, 7572, 2552, 449, 448, 8170, 7704, 1751, 8114, 1803, 8172, 5756, 7993, 7767, 472, 8173, 1750, 6089, 8086], "largest": 8174, "skip": false}
2026-05-17T14:03:09-05:00	INFO	hashstore	unclean shutdown detected: reconciling logs	{"satellite": "12EayRS2V1kEsWESU9QMRseFhdxYxKicsiFmxrsLZHeLUtdps3S", "store": "s1"}

Does this mean that the software is systematically checking my node’s status against the satellites’?

Anywho, initially I couldn’t access the dashboard, but after a few minutes of seeing these entries in the logs, the dashboard became available to access again. Status reads “Online” and QUIC is “OK.” I took a pretty big hit to online score, I’m in the low 80’s for all satellites now, but it does appear to be working at least.

Is there anything else I need to do to intervene at this point?

15

The hashstore backend has an integrated fsck for itself, so it may detect discrepancies between hashtbl and data logs and automatically correct them. It’s not related to communications with satellites, it’s a fully local check.
Your node should be online for the next 30 days to fully recover the online score. It may still fall for some time due to inertia, but then should start to recover.