USB died and new USB dying again! Moving to SSD/Flash!

[aex.perez]

Post Mortem, sort of

Well today things all running as they should after some settings I missed, which were dirving me nuts.

WiFi settings on the Router
Did not disable 802.11b
Had the wrong SSID shared key for the main/smartconnect SSID didn't notice as I use the SmartConnect SSID as the guest network and the guest networks 2.4Ghz and 5Ghz, as the actual internal SSID's

Syslog
Scribe, node logging, uiscribe now all setup again, working, and updated.
One node was logging correctly, the other was not (wouldn't even log logger) syslogd was failing to run correctly.
Message_loglevel and log_level were set wrong/too verbose on the nodes which I can only assume was part of the issue.
Both nodes had the same settings (for the stuff that matters).
Both filters on the router supporting the nodes were identical (except names, so each has there own log).

USB
Failure, not sure it was actually a failure of the USB as jffs, nvram and the USB were all impacted/corrupted and kept geting corrupted even after a restore (BACKUPMON)
Only after a Hardware reset (WPS reset) on the router, and starting from scratch (did the nodes as well to be safe) did it all start to operationally recover
What I thought was a failed USB, as nothing would load or run, and every script was resporting a locked file and claimed to be already running, is today working perfectly as much as you can say a USB works perfectly and has been for a few weeks now.
Router is on a UPS, fed by a good filtering power strip, nodes just plugged in but nothing else on the same circuits as those devices have had an issue, and really only the router went nuts.
Getting to SSD was a bonus, and has been very stable and the router just a bit more responsive
Having run @JGrana's MTD Checker tool on the router and nodes every few days just to be sure everything is stable in that regard, has shown it's all good across all three devices.

Lessons learned.
Even with backups, have a good retention schedule.
Even with backups have a secondary target, especially on the router. I have a secondary USB (FAT) targert that I can mount on the PC or Laptop and share to recover from there, if the NAS becomes unavailable when the router fails (especially if configured to support 802.3ad)
Document your setup parameters, because if all else fails and you have to start from scratch, you will inevitably forget something to get you back were you were.
And if/when you upgrade you router, and or move to a different code base (like 3004 to 3006) that will be very handy to at least get you started on on solid ground before making additional changes to take advantage of the new HW and or Software/Firmware.

VPNs
Mostly setup. Will likely sunset OpenVPN and stick to IPSEC (native to the Win11 Laptop, ease of use) and WireGuard (speed)

Back to low to very low maintenence mode now that its all done or at least normalcy is achived.

 

[Viktor Jaep]

Post Mortem, sort of

Well today things all running as they should after some settings I missed, which were dirving me nuts.

WiFi settings on the Router
Did not disable 802.11b
Had the wrong SSID shared key for the main/smartconnect SSID didn't notice as I use the SmartConnect SSID as the guest network and the guest networks 2.4Ghz and 5Ghz, as the actual internal SSID's

Syslog
Scribe, node logging, uiscribe now all setup again, working, and updated.
One node was logging correctly, the other was not (wouldn't even log logger) syslogd was failing to run correctly.
Message_loglevel and log_level were set wrong/too verbose on the nodes which I can only assume was part of the issue.
Both nodes had the same settings (for the stuff that matters).
Both filters on the router supporting the nodes were identical (except names, so each has there own log).

USB
Failure, not sure it was actually a failure of the USB as jffs, nvram and the USB were all impacted/corrupted and kept geting corrupted even after a restore (BACKUPMON)
Only after a Hardware reset (WPS reset) on the router, and starting from scratch (did the nodes as well to be safe) did it all start to operationally recover
What I thought was a failed USB, as nothing would load or run, and every script was resporting a locked file and claimed to be already running, is today working perfectly as much as you can say a USB works perfectly and has been for a few weeks now.
Router is on a UPS, fed by a good filtering power strip, nodes just plugged in but nothing else on the same circuits as those devices have had an issue, and really only the router went nuts.
Getting to SSD was a bonus, and has been very stable and the router just a bit more responsive
Having run @JGrana's MTD Checker tool on the router and nodes every few days just to be sure everything is stable in that regard, has shown it's all good across all three devices.

Lessons learned.
Even with backups, have a good retention schedule.
Even with backups have a secondary target, especially on the router. I have a secondary USB (FAT) targert that I can mount on the PC or Laptop and share to recover from there, if the NAS becomes unavailable when the router fails (especially if configured to support 802.3ad)
Document your setup parameters, because if all else fails and you have to start from scratch, you will inevitably forget something to get you back were you were.
And if/when you upgrade you router, and or move to a different code base (like 3004 to 3006) that will be very handy to at least get you started on on solid groud before making additional changes to take advantage of the new HW and or Software/Firmware.

VPNs
Mostly setup. Will likely sunset OpenVPN and stick to IPSEC (native to the Win11 Laptop, ease of use) and WireGuard (speed)

Back to low to very low maintenence mode now that its all done or at least normalcy is achived.

Thanks for all the great tips here, @aex.perez! Glad things have settled back down!

 

[aex.perez]

Post Mortem, sort of

Well today things all running as they should after some settings I missed, which were dirving me nuts.

WiFi settings on the Router
Did not disable 802.11b
Had the wrong SSID shared key for the main/smartconnect SSID didn't notice as I use the SmartConnect SSID as the guest network and the guest networks 2.4Ghz and 5Ghz, as the actual internal SSID's

Syslog
Scribe, node logging, uiscribe now all setup again, working, and updated.
One node was logging correctly, the other was not (wouldn't even log logger) syslogd was failing to run correctly.
Message_loglevel and log_level were set wrong/too verbose on the nodes which I can only assume was part of the issue.
Both nodes had the same settings (for the stuff that matters).
Both filters on the router supporting the nodes were identical (except names, so each has there own log).

USB
Failure, not sure it was actually a failure of the USB as jffs, nvram and the USB were all impacted/corrupted and kept geting corrupted even after a restore (BACKUPMON)
Only after a Hardware reset (WPS reset) on the router, and starting from scratch (did the nodes as well to be safe) did it all start to operationally recover
What I thought was a failed USB, as nothing would load or run, and every script was resporting a locked file and claimed to be already running, is today working perfectly as much as you can say a USB works perfectly and has been for a few weeks now.
Router is on a UPS, fed by a good filtering power strip, nodes just plugged in but nothing else on the same circuits as those devices have had an issue, and really only the router went nuts.
Getting to SSD was a bonus, and has been very stable and the router just a bit more responsive
Having run @JGrana's MTD Checker tool on the router and nodes every few days just to be sure everything is stable in that regard, has shown it's all good across all three devices.

Lessons learned.
Even with backups, have a good retention schedule.
Even with backups have a secondary target, especially on the router. I have a secondary USB (FAT) targert that I can mount on the PC or Laptop and share to recover from there, if the NAS becomes unavailable when the router fails (especially if configured to support 802.3ad)
Document your setup parameters, because if all else fails and you have to start from scratch, you will inevitably forget something to get you back were you were.
And if/when you upgrade you router, and or move to a different code base (like 3004 to 3006) that will be very handy to at least get you started on on solid ground before making additional changes to take advantage of the new HW and or Software/Firmware.

VPNs
Mostly setup. Will likely sunset OpenVPN and stick to IPSEC (native to the Win11 Laptop, ease of use) and WireGuard (speed)

Back to low to very low maintenence mode now that its all done or at least normalcy is achived.

Well, sort of.
One final thing I broke and now fixed by accident. Something I thought that I had fixed awhile ago but waas still broken.

My DirecTV streaming boxes, the Samsung Smart TV's, the Samsung SoundBar and a few other things, I set them back to using DHCP (LAN/DHCP Server, LAN/DNS Director set to Router & WAN Internet Connection/DoT) vs. having them with fixed IPs. In particular the DirecTV streaming boxes getting to picture from turning on, getting to Picture after changing channels, bringing up the guide, picking and changing a channel from the guide and getting to picture is almost instant, like it was before the router blew up so to speak. Set with Fixed IP I was having a very noticeable lag between these actions, it worked just slow enough to be a nuissance. If I were to give and example, when turning it on 5 seconds via DHCP and 15 seconds using a fixed IP. Even my Wife complained about the lag as she was commanding me to fix it!

On GigE wired, and 1Gb fiber still was slow with fixed vs. blazing set with DHCP. Gonna dig a little bit tomorrow before sharing on the DirecTVStream forum on Reddit to see if anyone else has noticed the same.

What I suspect is that some of these devices use hardcoded DNSes and IPV4 only when set to static. When set to DHCP, using what the router gives them, and IPv6 in addition to IPv4 instead.
Either way I'll take that win! I played with using other DNS, and no difference, using DHCP is what made the difference.

Now! I can get back to low, very low maintenence mode

 

[Tech9]

What is a very low maintenance mode for you with all the logging, Skynet, spdMerlin, connmon, scMerlin, vnstat? You mean checking things once a day or more often?

 

[aex.perez]

What is a very low maintenance mode for you with all the logging, Skynet, spdMerlin, connmon, scMerlin, vnstat? You mean checking things once a day or more often?

Ha!

There are some fishermen I knew, that could tell you what the weather was going to be like a day even a week in advance. They had an innate ability to read the flow of the wind, waves, basically the environment to get the natural flow of things and knew when that flow was off and what it meant. Logging is that for me.

With as many devices and interactions between them both local and across the WAN, logging doesn't point out excatly what is or what went wrong often, and can send you down a rabbit hole looking for things that aren't there if you're not careful. But it does give a sense of the ebb and flow of the network.

Logging didn't tell me how or why the Router bit it and I had to reset everything. I think it even sent me down the wrong path thinking only the USB failed. Though I will admit moving to SSD makes things work better. So I get your point, more than you know.

Just like the old men fishing can predict the weather, I knew that post would get a comment from you and it sort of makes my point.

Yes, these are consumer grade devices and stuff happens and logging may be exccesive if there was a huge amount of volume amd correlation I was trying to do. It's made less of an effort if monitoring just a few items. But in the stuff happens category, and at the risk of sounding like an informercial, there are two scripts in my signature that have helped me with that so when stuff happens I have an out and a way back to normalcy minimizing the amount of monitoring I do. At least until this last event, but I've taken additional steps since then to help mitigate.

Though I will admit I still do monitor because married to a passionate Latina with a passion bordering on addiction for her Novelas and a distaste for technology when it gets in her way, and is quite vocal about it. Yes, Monitoring but for my sanity and survival and collect happy life, happy wife rewards points for when things go wrong

Sidebar: I can cite several instances where logging didn't help but observation did. Going back in the day Novell 386 had the concept of memory pools, we were plagued with one particular server that like clock work, 10am and 2pm would ABEND or reboot. Would take out 100 users on it at a time at the Bank. Novell finally sent one guy to trouble shoot it, his tool was a black book of phone numbers to the engineers. Well long story short, 100 users on a server was the max and the way we were mapping dirves to directories would exceed a particular pool of memory causing the abend. No warnings or logs identified it but the ebb and flow of people logging on, particularly at 10am and 2pm (Bankers lunch ) did. One change to the Login Scripts to reduce the number of mapped drives was the answer, and nothing in the logs pointed to the problem.

 

[XIII]

Has anyone tried an external SSD in a USB flash driver cover?

Example: https://us.transcend-info.com/product/portable-ssd/esd310

Or can I better (less inference/heat?) use a regular SSD in a case with a USB cable?

(I might take advantage of Amazon Prime days and get a small discount on either purchase)

 

[Tech9]

I knew that post would get a comment from you

Yes, because we had this conversation before. You know I go straight - your family members periodically suffer the consecuences of your monitoring obsession. No one else in your housedold cares if the ISP maximum speed dropped from 940Mbps to 780Mbps for 1h 13min or if the latency increased from 10ms to 18ms between 6.42pm and 8.12pm. Everything will work exactly the same way and for sure with less issues and much less attention needed without USB drives, MerlinAU, BACKUPMON, Skynet, spdMerlin, scMerlin, scribe/uiscribe, ntpMerlin, connmon, vnstat. Your wife is very patient, actually. On top of everithing your network generates useless traffic to speedtest servers negatively affecting everyone else on your shared ISP line.

By the way, you have ntpMerlin twice in your signature, must like it a lot. Is the system time accurate? Are you sure?

 

[Tech9]

Or can I better (less inference/heat?) use a regular SSD in a case with a USB cable?

Better. Less heat transfer from/to the router. Perhaps cheaper too.

 

[XIII]

Or can I better (less inference/heat?) use a regular SSD in a case with a USB cable?

Considering this combo:

• SSD: https://www.verbatim-europe.com/en/internal-ssd/products/vi550-s3-ssd-256gb-49351
• Case: https://www.verbatim-europe.com/en/...tore-n-go-25-enclosure-kit-usb-32-gen-1-53106

Would that be a better option?

(The SSD "stick" is about €25 and this combo as well)

 

[Tech9]

I can only reccomend what I have tried myself and confirmed it's working with Asus routers - ORICO and UGREEN. Some USB3 cases have initialization issues at boot when used with routers. The drive inside brand/size/speed - doesn't matter. The cheapest one from a recognizable brand is good enough and will perhaps outlive any USB flash drive and the router itself.

USB recommendations for Diversion / Skynet?

Looking for recommendations for USB drives to be used for scripts / swap running Diversion / Skynet on an AX86U pro. Not interested in anything large-ish, as I only use this drive for running the amtm scripts and not for NAS in any way. Try to keep my io to a minimum since it pushes the CPU so...

www.snbforums.com