Alpine 3.8 and APScheduler

• Use alpine:3.8 as base image directly for latest package and OS updates;
• Update dependencies: gglsbl 1.4.14, and gunicorn 19.9.0;
• Replace use of crond with APScheduler Python package to run regular updates;
• Since crond is not longer used, no processes need to run as root and we can use USER Dockerfile directive to drop privileges.

Updated Dependencies

• Upgrade Flask to 1.0.2;
• Upgrade gglsbl to 1.4.11;
• Upgrade gunicorn to 19.8.1.

Updated Dependencies

• Upgraded to gglsbl-rest 1.4.10 and pid 2.2.0;
• Removed dead code and unnecessary WAL mode setting from update.py.

gglsbl-rest 1.5.2

Updated gglsbl to version 1.4.8 for fixes and improvements.

gglsbl-rest 1.5.1

Use alpine:3.7 as base image directly for latest package and OS updates.

ggsbl-rest v1.5.0

Run update.py and main gunicorn process as a regular gglsbl user instead of root for added security.

gglsbl-rest 1.4.0

• Use a single sqlite database in WAL mode (#10) for more simplicity, more cached data reuse, faster updates and reduced disk usage;
• Default value of WORKERS is now 8 per detected CPU core plus one.

gglsbl-rest 1.3.2

• Use Alpine 3.6 instead of 3.4, since this moves us from sqlite 3.13 to 3.20.1 and seems to solve afilipovich/gglsbl#28 with gglsbl 1.4.6;
• Use gglsbl 1.4.6 again.

ggsbl-rest v1.3.1

• Revert to gglsbl 1.4.5 to avoid exceptions on queries as per afilipovich/gglsbl#28;
• Delete SafeBrowsingList object after database update to ensure sqlite connection is closed before moving the database file.

gglsbl-rest 1.3.0

• Added environment variable KEEPALIVE to control persistent connection idle period;
• Added environment variable LIMIT_REQUEST_LINE to increase the default gunicorn maximum HTTP request line size, since we embed full URLs into our URI.