Category: Computing

Removing Greylisting

This is a follow up article to my post on setting up greylisting with postfix and Ubuntu. While I really like the idea behind greylisting, it was resulting in legitimate email not arriving.

Why it wasn’t working

  1. Frustrating delay for any password reset, at least a 300 second (5 minute) delay for legitimate email from a properly configured email server. This was a known issue, but still annoying.
  2. Recently (mid-August 2013) some hotmail servers (impacting at least @hotmail and @sympatico.ca email addresses) were returning bounces to the user instead of properly handling the greylisting. [Specific servers from my logs: snt0-omc1-s8.snt0.hotmail.com, blu0-omc1-s36.blu0.hotmail.com, dub0-omc1-s5.dub0.hotmail.com]
  3. Other individual servers that failed to handle greylisting correctly. One important example would be the Interac email transfer email servers (notify@payments.interac.ca).

The more I looked at the logs, the more legitimate email I found that wasn’t being delivered.

It was also difficult for people to understand. If a friend sends you an email, it is then bounced back to them due to their sending server not handling the greylisting correctly – they then send you a follow up email some 5+ minutes later, that email will come through just fine because it looks like a second attempt to deliver the original email. The broken sending server will then kept on the auto-whitelist for 30 days or so and greylisting will not be applied.

Any spammer smart or lucky enough to send via my backup mail server would miss the greylisting as that server didn’t use greylisting and email was always accepted from the trusted backup server.

Also, more spam software seems to be retrying now.

Tools. It wasn’t until very recently that I found the postgreyreport script, this is quite useful for generating a report on what was rejected by the greylisting your mail server is using. I’d recommend anyone using greylisting consider using this script to monitor what isn’t being delivered.

I’ll recommend an article I came across while investigating some of these issues. It is supportive of greylisting (which I can’t agree with now) but it does touch on some other techniques. It’s based on OpenBSD / spamd – something not (easily) available on Ubuntu.

Continue reading “Removing Greylisting”

Tiny Tiny RSS as alternative to Google Reader

Rich Site Summary (RSS) – also commonly called Really Simple Syndication attracted my attention early on in its rise to popularity. I liked that it gave me a way to keep up to date with a website / blog that posted material without having to visit that site to poll for new updates. Previous to RSS there were tools that alerted you to website updates, allowing you to keep tabs on many sites without the boring manual labour of visiting them all only to find nothing had changed.

Initially I used a stand alone desktop reader, I used several different ones as they evolved and even had a different set of feeds for work and home. Much after the initial launch of Google Reader did I switch to using it as it gave me a consistent experience across multiple machines – but I had to give up the ability to see intranet RSS feeds at work.

Once I had a smart phone, one of the first things I wanted to do was read my feeds on it. Enter NewsRob which provided fantastic support for Google Reader and had rock solid offline support.

It shouldn’t be news to anyone that Google Reader is closing down, so a couple of weeks ago I decided it was time to move to another solution so I could kick the tires before the shut down. I did look at a couple, but most of the alternatives simply want to have you move over to their free hosted solution. Certainly this is low effort, and probably the destination for most. In my search I came across Tiny Tiny RSS, a self hosted solution.

Short version of the story: it seems to fit my needs fairly well. The web client is good with some tweaks and the Android story is a bit weak relative to NewsRob (lacking good offline support) but I haven’t tried the official app yet.

If you want the long version – read on, I’ll cover installation and set up. Continue reading “Tiny Tiny RSS as alternative to Google Reader”

Fixing a fake USB flash drive

IMG_2836

 

A couple of years ago I picked up a conference give-away flash drive (4GB), which at the time seemed like a pretty nice freebie. The trouble was it only every liked to play nice with my Windows machine, Linux would refuse to mount it. The headline photo is the final product, I failed to take a before picture but the leather + snap case on this USB thumb drive was hideous anyway.

Turns out it was a fake, good thing it was free. Still in my typical fashion I didn’t want to just throw it away, heck I’m still carrying around the mysterious AMD 1GB key I got ages ago. So this bogus 4GB key sat in my work bag for a long while before I finally got to investigating it.

The very first thing I did was use the linux command lsusb, this helped me clue in that there was something wrong (fake) with the drive. I found a forum post that helped me get started down the right path. I got a copy of ChipGenius which told me the following:


Description: [I:]USB Mass Storage Device(Generic Flash Disk)

Device Type: Mass Storage Device

Protocal Version: USB 2.00
Current Speed: High Speed
Max Current: 100mA

USB Device ID: VID = 0011 PID = 7788
Serial Number: 874BE199

Device Vendor: Generic
Device Name: Mass Storage
Device Revision: 0103

Manufacturer: Generic
Product Model: Flash Disk
Product Revision: 8.00

Controller Vendor: Alcor Micro
Controller Part-Number: SC708(FC8708)/AU6987 - F/W EC23
Flash ID code: ADD5949A - Hynix H27UAG8T2BTR - 1CE/Single Channel [MLC-8K] -> Total Capacity = 2GB

So this felt like progress: it's 2GB and not 4GB as Windows seems to think. Still not bad for free. I then used my camera to get some close up shots of the naked circuit board to confirm the data that the ChipGenius tool dug out.

IMG_2810

IMG_2806

 

It was good to see that the values matched, this helped boost my confidence in the ChipGenius tool. I did find some references on the web claiming that sometimes ChipGenius is wrong, so it’s worth looking at the chips themselves.

Part of the output was also a link to a website, yet even with google translate the site left me guessing as to what I wanted to download – there were a lot of possible options. I choose one near the top “Series master, the Alcor MPtool AU6987T/6989 Yasukuni, production tools (2011.12.26.00)” as the title matches some of the data in the ChipGenius dump. In the end the stability of the site, language barrier and my inability to successfully download anything sent me off down other paths.

I then ended up searching on flashboot.ru with the controller chip number (FC8708) I was able to find and download (with a bit of google translate help) a tool that recognized the drive. The best way to find this tool is searching it’s name: FC_MpTool_FC8308_FC8508_FC8406_04.02.01.

FC MpTool

The user interface was mysterious, but clicking on the drive letter started a reformat.. which resulted in a 2GB flash drive. This newly formatted drive was quite happy under Linux.

In terms of performance, I benchmarked copying 7 x ~300MB video files (total 1.9GB) to the stick, this reported ~4.7MB/sec. There were certainly bursty updates in file file progress dialog in Ubuntu. This isn’t great, but again it was free and it works under Linux. I also tried zeroing the entire drive ($ sudo dd if=/dev/zero of=/dev/sdc) which reported: 2095054848 bytes (2.1 GB) copied, 335.285 s, 6.2 MB/s – again, not great but good enough.

A bit of heat shrink tubing applied to cover up the bare circuit board and I’ve got a hack worthy USB key.