1

Since hitting 7.1 our inventory has become almost unusable. I migrated the install to a new piece of hardware. 2x4corex3Ghz, 32GB mem with a mirror, 250GB spread across 4 drives RAID-0, 4 Nics teamed. No difference. Walked the whole ‘check this’ doc and cut everything back. As a last resort, I deleted all items with scan errors… or at least I tried, and managed to delete all devices. Great. Well, inventory runs wonderfully now that it’s empty…

So I added a new scan job. Starting small, to ensure I get everything in, and no errors. I scanned my switch management network first. SNMP picked up everything, then every device showed up in the ‘scan errors’ section because it couldn’t connect to SSH. I tested SSH, using putty. Credentials work fine. Checked firewall on server, it’s off. Tried many times to manually get the scan to work. No go. I get a ‘sockets error’ every time.

Any ideas??

(I should add, I tried scanning a small subnet of Workstations as well. They all errored out as well. Tested the WMI creds, and they are working correctly. This could be the same issue as above, but may be different. …working the switches first)

1 Spice up
2

OK I seem to have fixed this, sort of. Apparently the SNMP scan grabbed the hostname off the switch and stuck it in the configuration. My switch hostnames do not have registered DNS entries. When Spiceworks goes to grab the config, it apparently uses the hostname, not the IP address to connect to the device, which fails in my case. I added a DNS entry for the test switch, and it connected immediately and re-scanned.

I don’t really have any great desire to create DNS records for all of my switches. Not needed in my environment. Is there another way to make spiceworks use the IP address instead??

This is definitely different than the issue with the workstations, as all of those do have DNS entries.

3

Hi Ferule,

Spiceworks will use IP addresses when first scanning devices (during discovery scans). Granted, the scan may do some name resolution during discovery, but the device should still be discovered using the IP address.

If a device has already been discovered, Spiceworks will use the hostname during any group scans of that device.

If you delete one of the devices from your inventory and let Spiceworks discover it once more, is the device pulled in correctly?

4

Deleting, then re-scanning did not resolve the issue.

On all of these, I used the “Fix Issue” to enter the ssh creds. They all passed, and I clicked the “Rescan Now” button. 24 hours later, and they still have the error on them. Even though, I see they did connect (probably only that first time) and grab a configuration from the switch.

Today, when I was trying your delete/re-scan option, all of the remaining switches now seem fine. So there seems to be some disconnect on the scheduled scans and a manual.

I am stuck with a number of devices with “Scan Errors” that have no resolution, but won’t clear as there is no option to ignore… How are we supposed to clear all of these??

5

Hi Ferule,

It sounds like it would be a good idea for us to dig into your logs a bit to see what’s going on. Let me get a support ticket created for you and we can take a look.

6

I am also seeing this issue. (I am on latest version).

Is it a known bug?

Thanks,

Phil

7

Hi PhillipWray,

Could you send over an email to support@spiceworks.com so we can help out? There are a few known bugs related to scanning that could be affecting your scans.

8

Hi Jon,

Thanks will do - I’ll include the package logs as well.

Regards,

Phil

9

I think I’m having the same problem. Did you ever find a resolution? In my case the device is listed by hostname and says it is unable to identify the device. When I try to manually enter the credentials it gives the “Yep! Worked…” response but then fails to inventory. As soon as I create an entry in the HOSTS file to resolve the hostname to IP I am able to inventory the device. I have hundreds of devices like this. I have tried entering a handful into the HOSTS file and it worked for each one I have tried so far. I do not want to create a DNS entry for every device. I do not want to have to remember to add a new entry for every new device.

10

…correction on previous post… It will say “host unreachable or firewalled” when I try enter the credentials manually. Sorry I have been jumping all over the place trying to resolve all of the scan errors…

11

The solution for me was to move Spiceworks to it’s own dedicated VM. It was sharing with Kaspersky Anti Virus Server as this was somehow interrupting the network scans. (Thanks to Chad at Spiceworks for his support on this one.)