2014 year-end summary and 2015 work plan--incentive

Used to summarize, accustomed to the plan, no summary can not be scattered experience of the system, in the end is trees trees; no plan has no goal, no target people easily decadent, in the end is the benefits of the law paste couplets year is another year. For operation and maintenance work is more so, operation is self-driven, and development is demand-driven, which has a very big difference: operations involved in a very wide range of knowledge, specific work is also unusually dispersed, unpredictable unexpected events is to let you unprepared, did not summarize the work is very painful, always play the role of firefighters After several years of development, operation and maintenance of infrastructure systems, such as the majority of work through the script, the system to achieve automation, a large degree of liberation operation and maintenance of repetitive work, this time is very easy to do nothing, fear, this is deep experience, all plans for operation and maintenance is more important, plan to improve the self, planning optimization system ... have decided their own future!

The following summary of self-employment mail (only for the individual sensitive to replace the treatment), motivate themselves, welcome to shoot bricks.

Over the years, the 2014 will be the summary of the work and plan 2015 goals.

First, the operation and maintenance management work mainly completed:

(1) Platform MySQL slow query by day analysis push system on-line, will be based on the results of the classification of automatic mail to the relevant personnel;

(2) The study of Hadoop knowledge, after taking over Hadoop enriched the monitoring and warning project, increased the automatic capacity management and notification function, and a small number of optimization of the cluster;

(3) The creative realization of the domain-based single-public IP multi-site sites (HTTP&HTTPS) seven-layer NAT forwarding, end the intranet Web application testing needs memory ip+ port history (applied in Zhang * dob* and Zhang * Part of the test environment);

(4) Stone * * Room overall relocation to * * room and Pre-computer room site selection work;

(5) The online monitoring system adds support for multiple instances of Redis and proxy servers;

(6) * * Freight call center system learning and monitoring;

(7) Offline virtual machine system expansion and will * * Freight related virtual machine resource isolation distinction;

(8) Offline release system rewrite, combined with SVN has been perfectly implemented fully automated code release notification system.

Second, the specific business aspects are mainly completed:

(1) Upgrade and expansion of important server (increase the dual power supply, expand from the storage SSD disk, increase the backup device of the server/switch);

(2) Online business history data archiving and Cleanup (tra** Library, MongoDB Library);

(3) Platform image Server migration;

(4) Redis, Nginx Proxy, Haproxy single point of resolution (temporarily for cold switching on the line, Redis high-availability program has been completed and online stable operation for 2 months, 2015 early on-line);

(5) Sphinx Performance Optimization: By adding multi-instance and dedicated from the library to improve the performance of the word breaker, completely remove the Sphinx index Refresh on other business caused by the lag, blocking effect);

(6) New services in the online and offline environment deployment and automated release (MP service, Upload service, Impact *, dob**, etc.);

(7) Other daily operation and maintenance work (this aspect is a bit complicated, no longer listed).

Third, emergency response to deal with the main:

(1) * * Room and company network, server load, application itself, DB and other issues affecting normal access, operation and maintenance group timely troubleshooting, repair, follow-up, inform the relevant business parties;

(2) Some server hardware problems timely processing (such as the main library disk bad, Hadoop disk failure, Datanode motherboard failure can not boot, etc.);

(3) OpenSSL appears Heartbleed and bash vulnerability detection, the first time to troubleshoot, confirm, upgrade all related systems (see the Mail "about the company all Linux server Upgrade Bash 20140925").

Iv. Personal aspects:

(1) Acquired Red Hat's RHCA certification through study.

V. Major work plans for 2015 (if continued):

(1) Set up the offline account centralized management system (personal server account, Jira, SVN, Springboard and other unified management), the online fortress machine to manage the account is not necessary to redo;

(2) Building a bastion machine system (self-developed) for the unified portal of offline servers;

(3) Focus on the operation of the old server, and timely proposed hardware upgrade implementation plan;

(4) solve the problem of single point of service (see the Mail "Platform service single point failure carding 20141114");

(5) with the system becoming more and more complex, the offline Devbox has not been able to meet the needs of developers, bear * and Lee * In the early discussion of this matter, if feasible, we will cooperate with the bear * completed.

Eof

2014 year-end summary and 2015 work plan--incentive