Maibock

Orange-gold and a bit tipsy

Orange-gold and a bit tipsy

After eight consecutive beers that won’t see the light of day for quite awhile, I finally have another one to try. The Maibock is a very simple beer with only a handful of ingredients. A combination of Pilsner (8.75#) and Munich (4.5#) malt is offset by half an ounce of Magnum hops. The single 60 minute hop addition isn’t meant to add much in the way of flavor or aroma. All that Pilsner required a 90 minute boil and resulted in a 1.079 OG which is a bit above style. We fermented using German bock yeast (WLP833) but it spent a couple weeks at ale temps before there was enough space to get it into the lagering freezer.

Fermentation finished at 1.009 which is a tad dry for style. Over and under shooting the respective gravities translates into a good bit of booze: Beersmith estimates ABV at 9.3%. Alcohol is definitely prominent in the taste although it’s not overly harsh or solventy. There’s just not much for it to hide behind with minimal hop character and not a whole lot of sweetness. Fortunately the unconventional fermentation schedule didn’t result in any sulfide flavors or odors. With a normal spring I think I’d prefer the simple, clean malt flavor to be more prominent but with cool weather stubbornly sticking around that alcohol note elbowing its way to the front is less objectionable. The proper flavor is there it’s just playing second fiddle.

The Big Beers

CCCP Imperial Stout

CCCP Imperial Stout

The next foursome is another set of batches that will take awhile. Fermentation isn’t the issue here though; these boozy behemoths will just need a good long rest in the cellar to mellow out.

First was the only entry outside BJCP category nineteen – the Russian imperial stout. A big old dose of maris otter (15.25#) provided the base and black barley (1.25#) provided the stout. Special B (12 oz), Crystal 60 (6 oz), pale chocolate (6 oz), and regular chocolate (6 oz) rounded out the grains. Horizon (1.64 oz) at 60 minutes helps even out all that sweetness and East Kent Goldings (2 oz at 10 minutes, 1 oz at flame out) should lend flavor and aroma. With all those flavors present, clean and reliable WLP001 / California Ale was the only yeast. The post-boil OG checked in at 1.096 which was middling for this particular day. A sweet-ish FG 1.026 matched expectations.

Next we ventured west to England with an Old Ale. Maris otter (15.75#) remained the go-to European two row with crystal 60 (10 oz) and black patent malt (3 oz). The recipe called for black treacle (6 oz) which proved tough to find. We used blackstrap molasses as a hopefully appropriate substitute. Horizon (1.36 oz) at 60 minutes was again the bittering hop but with nothing else this may be a sweeter beer. Caramel and toffee notes are certainly appropriate for the style. The mildest OG of the day, as expected, was a mere 1.0.90. Mere. The WLP013 / London Ale yeast should match minor regional flavors. The FG after two weeks was 1.024.

Fermenter-ville

Fermenter-ville

Staying on the isles and getting even bigger was the English barleywine. Yet another bigger base of maris otter (17.5#) went with crystal 60 (0.5#) and crystal 120 (0.5#) for a simple but sizable grain bill. Sticking with the theme for the continent, the hops were Horizon (1.5 oz) at 60 minutes and East Kent Golding (0.6 oz at both 20 minutes and flame out). Unlike the previous two, this one boiled for 90 minutes to develop bigger gravity and more carmelization flavors. At the end of all that roiling we hit our highest starting gravity ever: 1.120. WLP013 / London Ale in a sizable starter managed to work through all that sugar without stalling out to finish at 1.024 FG.

Finally crossing to the other side of the Atlantic was the twenty-plus pound American barleywine. 18.5# of American two-row gets most of the way there. For flavor there was 0.75# each of crystal 15 and crystal 60 and 3oz each of special B and pale chocolate. For good/crazy measure 0.75# of cane sugar added a touch more gravity. The only real hophead of the quartet got 2 oz of Magnum at 60 minutes and a whole cavalcade of flavor and aroma at flame out: Centennial (1.5 oz), Amarillo (1.5 oz), and Chinook (1 oz). Trusty WLP 001 / California yeast has never failed in the past, but this one got a large starter just in case. Not to be outdone by its English counterpart, this one also clocked in at 1.120 OG. The finish was just a hair sweeter at 1.026.

These will rest in kegs long enough to carb and then move into bottles to grow old.

Sours Day

One challenge with trying to make every style of beer in 365 days is some styles take about 365 days to finish (or more even in the case of gueze). So best get these ones fermenting early on. Sours day was also the first attempt at four brews in one day. With two mash tuns and two kettles it’s fairly reasonable.

Mashing set up

Mashing set up

The first beer was a Flanders Red. The red had no definitive base malt with equal amounts pilsner and Vienna (4# each) and an array of specialty malts: 6 oz each of white wheat, caramel 60, special B, and aromatic malt along with 12 oz of Munich. Some serious evaporation during the boil gave an unexpectedly hefty OG of 1.074. East Kent Goldings (0.80 oz) at 60 minutes should contribute balance but not much flavor. WLP001 / California Ale yeast took care of primary fermentation.

The darker Flanders, Oud Bruin, was up next. The brown had a similar but not identical grain bill. Pilsner formed the base (8#) along with Munich (3.25#). Once again there were 6 oz each of white wheat, special B, and aromatic malt with a bigger dose of caramel 60 (10 oz) and a dash of black malt (1.6 oz). The second kettle didn’t boil as aggressively yielding a less concentrated 1.064 OG. The same Kent Goldings (1.20 oz) at 60 minutes ought to have the same effect and California Ale yeast got first crack at the available sugar.

Kettle central

Kettle central

The most challenging brew of the day was the straight (unblended) lambic. The style calls for “aged hops” which should act only as preservatives and add no bitterness whatsoever. No one sells that sort of thing, so based on internet recommendations we tried roasting whole leaf hops in a warm oven for hours. The end result was nearly aroma-less although the house was not odorless for quite some time after. The grain selections are just pilsner (6#) and flaked wheat (4#). The color was under style but 0.5 oz of midnight wheat fixed that. The four stage decoction mash took forever with a cooler mash tun and hot water additions to hit temperature targets. This style mash would really work better with a metal tun that could be heated directly. We added almost 3 oz of the oven-aged hops at 90 minutes and then 3 oz of malto-dextrine at the end of the boil. Like the Flanders ales, California Ale yeast handled initial fermentation. It’s reliable, clean, and nearly flavorless.

The last sour was a Belgian Specialty Ale which is a pretty wide open style. A good excuse to try a brettanomyces beer. This one had another simple grain bill with 8# of pilsner, 1.25# of caramel 60, and 13 oz of cane sugar for gravity. Unlock the other sours, this beer got a decent does of hops. Hallertauer (1.84 oz) at 60 minutes provided the bitterness then a whole lot of Styrian Goldings: 0.92 oz at 15 minutes, another once at flame out, and two more for dry hopping. This last one got a specialty yeast for flavor: WLP510 / Bastogne Belgian Ale.

All four beers spent six days in primary fermentation before being transferred to carboys for long term conditioning. The goal wasn’t necessarily to finish the primary stage but rather just to kick start the process. After transferring, the Belgian Specialty got a dose of Brettanomyces (WLP650) and the other three got a vial each of Belgian Sour Mix (WLP655). These are all much slower fermenting yeasts. If all went according to plan, the first round of yeasts left enough sugar for the second round to develop their unique flavors but not so much that it will take ages to finish. We’ll have a better idea in six months or so.

Irish Red and Dry Stout

100_0728

Irish Red Ale

 

Brew day number one on the way to eighty-two styles features two Irish beer styles. First is the Irish Red Ale. The base malt was Maris Otter (10#) with Cara 40 (6 oz) and Cara 120 (6 oz) for sweetness and black barley (4 oz) to bring the color into the correct range. With a pretty standard mash OG came in at 1.054. Yeast was WLP004 / Irish Ale and the only hop addition was 1.25 oz of East Kent Goldings at 60 minutes. Fermentation finished at 1.014. The IBUs overshot the style by a hair but otherwise the numbers all came in where they belong. This isn’t one of my favorite styles but I do think it’s a successful beer. Sweetness is the predominant flavor but it’s quite clean with enough bitterness to avoid being cloying. The color is definitely coppery when it catches the light.

100_0731

Dry Stout

The other beer from the first day was a Dry Stout. The base malt was also Maris Otter (5.75#) along with lots of barley – both flaked (2#) and black (0.75#). There was also de-bittered black (0.25#) to get that great motor oil color. This mash included a fifteen minute protein rest because of all that barley. OG was 1.048. Once again the yeast was WLP004 and the only hop addition was East Kent Goldings at 60 minutes but this beer needed 1.75oz of hops. Fermentation stopped at an appropriately dry FG of 1.010. The stout hit all its numbers but the ABV was right on the line. Dry stout is a style I like and I like this particular example quite a lot. It’s big and roasty without being overly bitter. Despite the dryness it has plenty of body and of course the perfect jet black color. Granted, black is probably the easiest color to create.

Securing this with SSL

Time to apply some proper certificates via good ‘ol StartSSL. After validating the domain via email, I need an SSL/TLS certificate. I generate my own signing request:

# cd /etc/ssl/
# openssl req -new -newkey rsa:2048 -nodes -keyout outofadjectives.org.key -out outofadjectives.org.csr

Just copy/paste the CSR and finish up the StartSSL wizard and copy/paste the resulting key into a .crt file. It corresponds with the previously generated .key.

The nginx instructions are the most straightforward. The unified certificate format makes configuration for everything else simpler since there are no chain files to deal with. The password step might not be necessary depending whether you used a password with the CSR.

To use the certificate in Dovecot, edit /etc/dovecot/conf.d/10-ssl.conf so ssl_cert points at the unified certificate file and ssl_key points at the .key file. Then restart the service.

For Postfix, edit /etc/postfix/main.cf so that smtpd_tls_cert_file and smtp_tls_cert_file both point at the unified certificate file and smtpd_tls_key_file and smtp_tls_key_file both point at the .key file. The “smtp_” settings may not exist by default. Again, restart the service.

PDO and only PDO

Life on EC2 can be a bit cramped (or alternately, expensive) in the way of memory. I wanted to slim down my memory footprint for PHP by disabling extra extensions. The biggest culprit in terms of redundancy was MySQL with mysql, mysqli, and pdo_mysql. Using just the mysql extension would be the easy route, but not the future-proofing route. I picked pdo_mysql for a slight edge in flexibility – even if SQL-agnostic applications are a fantasy.

First up for compatibility checking was roundcube. The version available via apt can only use the mysql extension, but that’s a fairly old version and development on the project is still active. Roundcube has moved over to github from SVN which is a nice change. The development version is built entirely around PDO and works quite nicely so far. Upgrading was as simple as copying a few configuration directives from the old version to the new one.

Next was WordPress. The base version still requires mysql and it looks like that will be the case for quite awhile due to compatibility and availability concerns. But some WordPress developers have contributed a set of preliminary patches to allow mysqli and pdo_mysql usage. After a little tinkering, I applied those patches to the latest development release and the blog now runs on PDO too (knock on wood). I made my own fork on github with a PDO branch since tracking the upstream version is going to be a bit more complicated with local changes to maintain.