Administering Bugzilla Or, I just got this cool thing installed. Now what the heck do I do with it? So you followed to the letter, and logged into Bugzilla for the very first time with your super-duper god account. You sit, contentedly staring at the Bugzilla Query Screen, the worst of the whole mad business of installing this terrific program behind you. It seems, though, you have nothing yet to query! Your first act of business should be to setup the operating parameters for Bugzilla so you can get busy getting data into your bug tracker.
Post-Installation Checklist After installation, follow the checklist below to help ensure that you have a successful installation. If you do not see a recommended setting for a parameter, consider leaving it at the default while you perform your initial tests on your Bugzilla setup. checklist Bring up editparams.cgi in your web browser. This should be available as the edit parameters link from any Bugzilla screen once you have logged in. The maintainer is the email address of the person responsible for maintaining this Bugzilla installation. The maintainer need not be a valid Bugzilla user. Error pages, error emails, and administrative mail will be sent with the maintainer as the return email address. Set maintainer to your email address. This allows Bugzilla's error messages to display your email address and allow people to contact you for help. The urlbase parameter defines the fully qualified domain name and web server path to your Bugzilla installation. For example, if your bugzilla query page is http://www.foo.com/bugzilla/query.cgi, set your urlbase is http://www.foo.com/bugzilla/. usebuggroups dictates whether or not to implement group-based security for Bugzilla. If set, Bugzilla bugs can have an associated groupmask defining which groups of users are allowed to see and edit the bug. Set "usebuggroups" to "on" only if you may wish to restrict access to products. I suggest leaving this parameter off while initially testing your Bugzilla. usebuggroupsentry, when set to on, requires that all bugs have an associated groupmask when submitted. This parameter is made for those installations where product isolation is a necessity. Set "usebuggroupsentry" to "on" if you absolutely need to restrict access to bugs from the moment they are submitted through resolution. Once again, if you are simply testing your installation, I suggest against turning this parameter on; the strict security checking may stop you from being able to modify your new entries. You run into an interesting problem when Bugzilla reaches a high level of continuous activity. MySQL supports only table-level write locking. What this means is that if someone needs to make a change to a bug, they will lock the entire table until the operation is complete. Locking for write also blocks reads until the write is complete. The shadowdb parameter was designed to get around this limitation. While only a single user is allowed to write to a table at a time, reads can continue unimpeded on a read-only shadow copy of the database. Although your database size will double, a shadow database can cause an enormous performance improvement when implemented on extremely high-traffic Bugzilla databases. Set "shadowdb" to "bug_shadowdb" if you will be running a *very* large installation of Bugzilla. The shadow database enables many simultaneous users to read and write to the database without interfering with one another. Enabling "shadowdb" can adversely affect the stability of your installation of Bugzilla. You should regularly check that your database is in sync. It is often advisable to force a shadow database sync nightly via cron. Once again, in testing you should avoid this option -- use it if or when you need to use it, and have repeatedly run into the problem it was designed to solve -- very long wait times while attempting to commit a change to the database. Mozilla.org began needing shadowdb when they reached around 40,000 Bugzilla users with several hundred Bugzilla bug changes and comments per day. If you use the "shadowdb" option, it is only natural that you should turn the "queryagainstshadowdb" option "On" as well. Otherwise you are replicating data into a shadow database for no reason! headerhtml, footerhtml, errorhtml, bannerhtml, and blurbhtml are all templates which control display of headers, footers, errors, banners, and additional data. We could go into some detail regarding the usage of these, but it is really best just to monkey around with them a bit to see what they do. I strongly recommend you copy your data/params file somewhere safe before playing with these values, though. If they are changed dramatically, it may make it impossible for you to display Bugzilla pages to fix the problem until you have restored your data/params file. If you have custom logos or HTML you must put in place to fit within your site design guidelines, place the code in the "headerhtml", "footerhtml", "errorhtml", "bannerhtml", or "blurbhtml" text boxes. The "headerhtml" text box is the HTML printed out before any other code on the page, except the CONTENT-TYPE header sent by the Bugzilla engine. If you have a special banner, put the code for it in "bannerhtml". You may want to leave these settings at the defaults initially. passwordmail is rather simple. Every time a user creates an account, the text of this parameter is read as the text to send to the new user along with their password message. Add any text you wish to the "passwordmail" parameter box. For instance, many people choose to use this box to give a quick training blurb about how to use Bugzilla at your site. useqacontact allows you to define an email address for each component, in addition to that of the default owner, who will be sent carbon copies of incoming bugs. The critical difference between a QA Contact and an Owner is that the QA Contact follows the component. If you reassign a bug from component A to component B, the QA Contact for that bug will change with the reassignment, regardless of owner. usestatuswhiteboard defines whether you wish to have a free-form, overwritable field associated with each bug. The advantage of the Status Whiteboard is that it can be deleted or modified with ease, and provides an easily-searchable field for indexing some bugs that have some trait in common. Many people will put help wanted, stalled, or waiting on reply from somebody messages into the Status Whiteboard field so those who peruse the bugs are aware of their status even more than that which can be indicated by the Resolution fields. Do you want to use the QA Contact ("useqacontact") and status whiteboard ("usestatuswhiteboard") fields? These fields are useful because they allow for more flexibility, particularly when you have an existing Quality Assurance and/or Release Engineering team, but they may not be needed for many smaller installations. Set "whinedays" to the amount of days you want to let bugs go in the "New" or "Reopened" state before notifying people they have untouched new bugs. If you do not plan to use this feature, simply do not set up the whining cron job described in the installation instructions, or set this value to "0" (never whine). commenton fields allow you to dictate what changes can pass without comment, and which must have a comment from the person who changed them. Often, administrators will allow users to add themselves to the CC list, accept bugs, or change the Status Whiteboard without adding a comment as to their reasons for the change, yet require that most other changes come with an explanation. Set the "commenton" options according to your site policy. It is a wise idea to require comments when users resolve, reassign, or reopen bugs at the very least. It is generally far better to require a developer comment when resolving bugs than not. Few things are more annoying to bug database users than having a developer mark a bug "fixed" without any comment as to what the fix was (or even that it was truly fixed!) The supportwatchers option can be an exceptionally powerful tool in the hands of a power Bugzilla user. By enabling this option, you allow users to receive email updates whenever other users receive email updates. This is, of course, subject to the groupset restrictions on the bug; if the watcher would not normally be allowed to view a bug, the watcher cannot get around the system by setting herself up to watch the bugs of someone with bugs outside her privileges. She would still only receive email updates for those bugs she could normally view. For Bugzilla sites which require strong inter-Product security to prevent snooping, watchers are not a good idea. However, for most sites you should set supportwatchers to "On". This feature is helpful for team leads to monitor progress in their respective areas, and can offer many other benefits, such as allowing a developer to pick up a former engineer's bugs without requiring her to change all the information in the bug.
User Administration User administration is one of the easiest parts of Bugzilla. Keeping it from getting out of hand, however, can become a challenge.
Creating the Default User When you first run checksetup.pl after installing Bugzilla, it will prompt you for the administrative username (email address) and password for this "super user". If for some reason you were to delete the "super user" account, re-running checksetup.pl will again prompt you for this username and password. If you wish to add more administrative users, you must use the MySQL interface. Run "mysql" from the command line, and use these commands ("mysql>" denotes the mysql prompt, not something you should type in): mysql> use bugs; mysql> update profiles set groupset=0x7ffffffffffffff where login_name = "(user's login name)"; Yes, that is fourteen f's. A whole lot of f-ing going on if you want to create a new administator.
Managing Other Users
Logging In Open the index.html page for your Bugzilla installation in your browser window. Click the "Query Existing Bug Reports" link. Click the "Log In" link at the foot of the page. Type your email address, and the password which was emailed to you when you created your Bugzilla account, into the spaces provided. Congratulations, you are logged in!
Creating new users Your users can create their own user accounts by clicking the "New Account" link at the bottom of each page. However, should you desire to create user accounts ahead of time, here is how you do it. After logging in, click the "Users" link at the footer of the query page. To see a specific user, type a portion of their login name in the box provided and click "submit". To see all users, simply click the "submit" button. You must click "submit" here to be able to add a new user. More functionality is available via the list on the right-hand side of the text entry box. You can match what you type as a case-insensitive substring (the default) of all users on your system, a case-sensitive regular expression (please see the man regexp manual page for details on regular expression syntax), or a reverse regular expression match, where every user name which does NOT match the regular expression is selected. Click the "Add New User" link at the bottom of the user list Fill out the form presented. This page is self-explanatory. When done, click "submit". Adding a user this way will not send an email informing them of their username and password. While useful for creating dummy accounts (watchers which shuttle mail to another system, for instance, or email addresses which are a mailing list), in general it is preferable to log out and use the New Account button to create users, as it will pre-populate all the required fields and also notify the user of her account name and password.
Disabling Users I bet you noticed that big "Disabled Text" entry box available from the "Add New User" screen, when you edit an account? By entering any text in this box and selecting "submit", you have prevented the user from using Bugzilla via the web interface. Your explanation, written in this text box, will be presented to the user the next time she attempts to use the system. Don't disable your own administrative account, or you will hate life! At this time, Disabled Text does not prevent a user from using the email interface. If you have the email interface enabled, they can still continue to submit bugs and comments that way. We need a patch to fix this.
Modifying Users Here I will attempt to describe the function of each option on the Edit User screen. Login Name: This is generally the user's email address. However, if you have edited your system parameters, this may just be the user's login name or some other identifier. For compatability reasons, you should probably stick with email addresses as user login names. It will make your life easier. Real Name: Duh! Password: You can change the user password here. It is normal to only see asterisks. Email Notification: You may choose from one of three options: All qualifying bugs except those which I change: The user will be notified of any change to any bug for which she is the reporter, assignee, QA Contact, CC recipient, or "watcher". Only those bugs which I am listed on the CC line: The user will not be notified of changes to bugs where she is the assignee, reporter, or QA Contact, but will receive them if she is on the CC list. She will still receive whining cron emails if you set up the "whinemail" feature. All Qualifying Bugs: This user is a glutton for punishment. If her name is in the reporter, QA Contact, CC, assignee, or is a "watcher", she will get email updates regarding the bug. Disable Text: If you type anything in this box, including just a space, the user account is disabled from making any changes to bugs via the web interface, and what you type in this box is presented as the reason. Don't disable the administrator account! As of this writing, the user can still submit bugs via the e-mail gateway, if you set it up, despite the disabled text field. The e-mail gateway should not be enabled for secure installations of Bugzilla. CanConfirm: This field is only used if you have enabled "unconfirmed" status in your parameters screen. If you enable this for a user, that user can then move bugs from "Unconfirmed" to "Confirmed" status (e.g.: "New" status). Be judicious about allowing users to turn this bit on for other users. Creategroups: This option will allow a user to create and destroy groups in Bugzilla. Unless you are using the Bugzilla GroupSentry security option "usebuggroupsentry" in your parameters, this setting has no effect. Editbugs: Unless a user has this bit set, they can only edit those bugs for which they are the assignee or the reporter. Leaving this option unchecked does not prevent users from adding comments to a bug! They simply cannot change a bug priority, severity, etc. unless they are the assignee or reporter. Editcomponents: This flag allows a user to create new products and components, as well as modify and destroy those that have no bugs associated with them. If a product or component has bugs associated with it, those bugs must be moved to a different product or component before Bugzilla will allow them to be destroyed. The name of a product or component can be changed without affecting the associated bugs, but it tends to annoy the hell out of your users when these change a lot. Editkeywords: If you use Bugzilla's keyword functionality, enabling this feature allows a user can create and destroy keywords. As always, the keywords for existing bugs containing the keyword the user wishes to destroy must be changed before Bugzilla will allow it to die. You must be very careful about creating too many new keywords if you run a very large Bugzilla installation; keywords are global variables across products, and you can often run into a phenomenon called "keyword bloat". This confuses users, and then the feature goes unused. Editusers: This flag allows a user do what you're doing right now: edit other users. This will allow those with the right to do so to remove administrator privileges from other users or grant them to themselves. Enable with care. PRODUCT: PRODUCT bugs access. This allows an administrator, with product-level granularity, to specify in which products a user can edit bugs. The user must still have the "editbugs" privelege to edit bugs in this area; this simply restricts them from even seeing bugs outside these boundaries if the administrator has enabled the group sentry parameter "usebuggroupsentry". Unless you are using bug groups, this option has no effect.
Product, Component, Milestone, and Version Administration Dear Lord, we have to get our users to do WHAT?
Products Formerly, and in some spots still, called "Programs" Products are the broadest category in Bugzilla, and you should have the least of these. If your company makes computer games, you should have one product per game, and possibly a few special products (website, meetings...) A Product (formerly called "Program", and still referred to that way in some portions of the source code) controls some very important functions. The number of "votes" available for users to vote for the most important bugs is set per-product, as is the number of votes required to move a bug automatically from the UNCONFIRMED status to the NEW status. One can close a Product for further bug entry and define various Versions available from the Edit product screen. To create a new product: Select "components" from the yellow footer It may seem counterintuitive to click "components" when you want to edit the properties associated with Products. This is one of a long list of things we want in Bugzilla 3.0... Select the "Add" link to the right of "Add a new product". Enter the name of the product and a description. The Description field is free-form. Don't worry about the "Closed for bug entry", "Maximum Votes per person", "Maximum votes a person can put on a single bug", "Number of votes a bug in this Product needs to automatically get out of the UNCOMFIRMED state", and "Version" options yet. We'll cover those in a few moments.
Components Components are subsections of a Product. Creating some Components The computer game you are designing may have a "UI" component, an "API" component, a "Sound System" component, and a "Plugins" component, each overseen by a different programmer. It often makes sense to divide Components in Bugzilla according to the natural divisions of responsibility within your Product or company. Each component has a owner and (if you turned it on in the parameters), a QA Contact. The owner should be the primary person who fixes bugs in that component. The QA Contact should be the person who will ensure these bugs are completely fixed. The Owner, QA Contact, and Reporter will get email when new bugs are created in this Component and when these bugs change. Default Owner and Default QA Contact fields only dictate the default assignments; the Owner and QA Contact fields in a bug are otherwise unrelated to the Component. To create a new Component: Select the "Edit components" link from the "Edit product" page Select the "Add" link to the right of the "Add a new component" text on the "Select Component" page. Fill out the "Component" field, a short "Description", and the "Initial Owner". The Component and Description fields are free-form; the "Initial Owner" field must be that of a user ID already existing in the database. If the initial owner does not exist, Bugzilla will refuse to create the component. Is your "Default Owner" a user who is not yet in the database? No problem. Select the "Log out" link on the footer of the page. Select the "New Account" link on the footer of the "Relogin" page Type in the email address of the default owner you want to create in the "E-mail address" field, and her full name in the "Real name" field, then select the "Submit Query" button. Now select "Log in" again, type in your login information, and you can modify the product to use the Default Owner information you require. Either Edit more components or return to the Bugzilla Query Page. To return to the Product you were editing, you must select the Components link as before.
Versions Versions are the revisions of the product, such as "Flinders 3.1", "Flinders 95", and "Flinders 2000". Using Versions helps you isolate code changes and are an aid in reporting. Common Use of Versions A user reports a bug against Version "Beta 2.0" of your product. The current Version of your software is "Release Candidate 1", and no longer has the bug. This will help you triage and classify bugs according to their relevance. It is also possible people may report bugs against bleeding-edge beta versions that are not evident in older versions of the software. This can help isolate code changes that caused the bug A Different Use of Versions This field has been used to good effect by an online service provider in a slightly different way. They had three versions of the product: "Production", "QA", and "Dev". Although it may be the same product, a bug in the development environment is not normally as critical as a Production bug, nor does it need to be reported publicly. When used in conjunction with Target Milestones, one can easily specify the environment where a bug can be reproduced, and the Milestone by which it will be fixed. To create and edit Versions: From the "Edit product" screen, select "Edit Versions" You will notice that the product already has the default version "undefined". If your product doesn't use version numbers, you may want to leave this as it is or edit it so that it is "---". You can then go back to the edit versions page and add new versions to your product. Otherwise, click the "Add" button to the right of the "Add a new version" text. Enter the name of the Version. This can be free-form characters up to the limit of the text box. Then select the "Add" button. At this point you can select "Edit" to edit more Versions, or return to the "Query" page, from which you can navigate back to the product through the "components" link at the foot of the Query page.
Milestones Milestones are "targets" that you plan to get a bug fixed by. For example, you have a bug that you plan to fix for your 3.0 release, it would be assigned the milestone of 3.0. Or, you have a bug that you plan to fix for 2.8, this would have a milestone of 2.8. Milestone options will only appear for a Product if you turned the "usetargetmilestone" field in the "Edit Parameters" screen "On". To create new Milestones, set Default Milestones, and set Milestone URL: Select "edit milestones" Select "Add" to the right of the "Add a new milestone" text Enter the name of the Milestone in the "Milestone" field. You can optionally set the "Sortkey", which is a positive or negative number (-255 to 255) that defines where in the list this particular milestone appears. Select "Add". Using SortKey with Target Milestone Let's say you create a target milestone called "Release 1.0", with Sortkey set to "0". Later, you realize that you will have a public beta, called "Beta1". You can create a Milestone called "Beta1", with a Sortkey of "-1" in order to ensure people will see the Target Milestone of "Beta1" earlier on the list than "Release 1.0" If you want to add more milestones, select the "Edit" link. If you don't, well shoot, you have to go back to the "query" page and select "components" again, and make your way back to the Product you were editing. This is another in the list of unusual user interface decisions that we'd like to get cleaned up. Shouldn't there be a link to the effect of "edit the Product I was editing when I ended up here"? In any case, clicking "components" in the footer takes you back to the "Select product" screen, from which you can begin editing your product again. From the Edit product screen again (once you've made your way back), enter the URL for a description of what your milestones are for this product in the "Milestone URL" field. It should be of the format "http://www.foo.com/bugzilla/product_milestones.html" Some common uses of this field include product descriptions, product roadmaps, and of course a simple description of the meaning of each milestone. If you're using Target Milestones, the "Default Milestone" field must have some kind of entry. If you really don't care if people set coherent Target Milestones, simply leave this at the default, "---". However, controlling and regularly updating the Default Milestone field is a powerful tool when reporting the status of projects. Select the "Update" button when you are done.
Voting The concept of "voting" is a poorly understood, yet powerful feature for the management of open-source projects. Each user is assigned so many Votes per product, which they can freely reassign (or assign multiple votes to a single bug). This allows developers to gauge user need for a particular enhancement or bugfix. By allowing bugs with a certain number of votes to automatically move from "UNCONFIRMED" to "NEW", users of the bug system can help high-priority bugs garner attention so they don't sit for a long time awaiting triage. The daunting challenge of Votes is deciding where you draw the line for a "vocal majority". If you only have a user base of 100 users, setting a low threshold for bugs to move from UNCONFIRMED to NEW makes sense. As the Bugzilla user base expands, however, these thresholds must be re-evaluated. You should gauge whether this feature is worth the time and close monitoring involved, and perhaps forego implementation until you have a critical mass of users who demand it. To modify Voting settings: Navigate to the "Edit product" screen for the Product you wish to modify Set "Maximum Votes per person" to your calculated value. Setting this field to "0" disables voting. Set "Maximum Votes a person can put on a single bug" to your calculated value. It should probably be some number lower than the "Maximum votes per person". Setting this field to "0" disables voting, but leaves the voting options open to the user. This is confusing. Set "Number of votes a bug in this product needs to automatically get out of the UNCONFIRMED state" to your calculated number. Setting this field to "0" disables the automatic move of bugs from UNCONFIRMED to NEW. Some people advocate leaving this at "0", but of what use are Votes if your Bugzilla user base is unable to affect which bugs appear on Development radar? You should probably set this number to higher than a small coalition of Bugzilla users can influence it. Most sites use this as a "referendum" mechanism -- if users are able to vote a bug out of UNCONFIRMED, it is a really bad bug! Once you have adjusted the values to your preference, select the "Update" button.
Groups and Group Security Groups can be very useful in bugzilla, because they allow users to isolate bugs or products that should only be seen by certain people. Groups can also be a complicated minefield of interdependencies and weirdness if mismanaged. When to Use Group Security Many Bugzilla sites isolate "Security-related" bugs from all other bugs. This way, they can have a fix ready before the security vulnerability is announced to the world. You can create a "Security" product which, by default, has no members, and only add members to the group (in their individual User page, as described under User Administration) who should have priveleged access to "Security" bugs. Alternately, you may create a Group independently of any Product, and change the Group mask on individual bugs to restrict access to members only of certain Groups. Groups only work if you enable the "usebuggroups" paramater. In addition, if the "usebuggroupsentry" parameter is "On", one can restrict access to products by groups, so that only members of a product group are able to view bugs within that product. Group security in Bugzilla can be divided into two categories: Generic and Product-Based. Groups in Bugzilla are a complicated beast that evolved out of very simple user permission bitmasks, apparently itself derived from common concepts in UNIX access controls. A "bitmask" is a fixed-length number whose value can describe one, and only one, set of states. For instance, UNIX file permissions are assigned bitmask values: "execute" has a value of 1, "write" has a value of 2, and "read" has a value of 4. Add them together, and a file can be read, written to, and executed if it has a bitmask of "7". (This is a simplified example -- anybody who knows UNIX security knows there is much more to it than this. Please bear with me for the purpose of this note.) The only way a bitmask scheme can work is by doubling the bit count for each value. Thus if UNIX wanted to offer another file permission, the next would have to be a value of 8, then the next 16, the next 32, etc. Similarly, Bugzilla offers a bitmask to define group permissions, with an internal limit of 64. Several are already occupied by built-in permissions. The way around this limitation is to avoid assigning groups to products if you have many products, avoid bloating of group lists, and religiously prune irrelevant groups. In reality, most installations of Bugzilla support far fewer than 64 groups, so this limitation has not hit for most sites, but it is on the table to be revised for Bugzilla 3.0 because it interferes with the security schemes of some administrators. To enable Generic Group Security ("usebuggroups"): Turn "On" "usebuggroups" in the "Edit Parameters" screen. You will generally have no groups set up. Select the "groups" link in the footer. Take a moment to understand the instructions on the "Edit Groups" screen. Once you feel confident you understand what is expected of you, select the "Add Group" link. Fill out the "New Name" (remember, no spaces!), "New Description", and "New User RegExp" fields. "New User RegExp" allows you to automatically place all users who fulfill the Regular Expression into the new group. Creating a New Group I created a group called DefaultGroup with a description of This is simply a group to play with, and a New User RegExp of .*@mydomain.tld. This new group automatically includes all Bugzilla users with "@mydomain.tld" at the end of their user id. When I finished, my new group was assigned bit #128. When you have finished, select the Add button. To enable Product-Based Group Security (usebuggroupsentry): Don't forget that you only have 64 groups masks available, total, for your installation of Bugzilla! If you plan on having more than 50 products in your individual Bugzilla installation, and require group security for your products, you should consider either running multiple Bugzillas or using Generic Group Security instead of Product-Based ("usebuggroupsentry") Group Security. Turn "On" "usebuggroups" and "usebuggroupsentry" in the "Edit Parameters" screen. "usebuggroupsentry" has the capacity to prevent the administrative user from directly altering bugs because of conflicting group permissions. If you plan on using "usebuggroupsentry", you should plan on restricting administrative account usage to administrative duties only. In other words, manage bugs with an unpriveleged user account, and manage users, groups, Products, etc. with the administrative account. You will generally have no Groups set up, unless you enabled "usebuggroupsentry" prior to creating any Products. To create "Generic Group Security" groups, follow the instructions given above. To create Product-Based Group security, simply follow the instructions for creating a new Product. If you need to add users to these new groups as you create them, you will find the option to add them to the group available under the "Edit User" screens. You may find this example illustrative for how bug groups work. Bugzilla Groups Bugzilla Groups example ----------------------- For this example, let us suppose we have four groups, call them Group1, Group2, Group3, and Group4. We have 5 users, User1, User2, User3, User4, User5. We have 8 bugs, Bug1, ..., Bug8. Group membership is defined by this chart: (X denotes that user is in that group.) (I apologize for the nasty formatting of this table. Try viewing it in a text-based browser or something for now. -MPB) G G G G r r r r o o o o u u u u p p p p 1 2 3 4 +-+-+-+-+ User1|X| | | | +-+-+-+-+ User2| |X| | | +-+-+-+-+ User3|X| |X| | +-+-+-+-+ User4|X|X|X| | +-+-+-+-+ User5| | | | | +-+-+-+-+ Bug restrictions are defined by this chart: (X denotes that bug is restricted to that group.) G G G G r r r r o o o o u u u u p p p p 1 2 3 4 +-+-+-+-+ Bug1| | | | | +-+-+-+-+ Bug2| |X| | | +-+-+-+-+ Bug3| | |X| | +-+-+-+-+ Bug4| | | |X| +-+-+-+-+ Bug5|X|X| | | +-+-+-+-+ Bug6|X| |X| | +-+-+-+-+ Bug7|X|X|X| | +-+-+-+-+ Bug8|X|X|X|X| +-+-+-+-+ Who can see each bug? Bug1 has no group restrictions. Therefore, Bug1 can be seen by any user, whatever their group membership. This is going to be the only bug that User5 can see, because User5 isn't in any groups. Bug2 can be seen by anyone in Group2, that is User2 and User4. Bug3 can be seen by anyone in Group3, that is User3 and User4. Bug4 can be seen by anyone in Group4. Nobody is in Group4, so none of these users can see Bug4. Bug5 can be seen by anyone who is in _both_ Group1 and Group2. This is only User4. User1 cannot see it because he is not in Group2, and User2 cannot see it because she is not in Group1. Bug6 can be seen by anyone who is in both Group1 and Group3. This would include User3 and User4. Similar to Bug5, User1 cannot see Bug6 because he is not in Group3. Bug7 can be seen by anyone who is in Group1, Group2, and Group3. This is only User4. All of the others are missing at least one of those group privileges, and thus cannot see the bug. Bug8 can be seen by anyone who is in Group1, Group2, Group3, and Group4. There is nobody in all four of these groups, so nobody can see Bug8. It doesn't matter that User4 is in Group1, Group2, and Group3, since he isn't in Group4.
Bugzilla Security Putting your money in a wall safe is better protection than depending on the fact that no one knows that you hide your money in a mayonnaise jar in your fridge. Poorly-configured MySQL, Bugzilla, and FTP installations have given attackers full access to systems in the past. Please take these guidelines seriously, even for Bugzilla machines hidden away behind your firewall. 80% of all computer trespassers are insiders, not anonymous crackers. Secure your installation. These instructions must, of necessity, be somewhat vague since Bugzilla runs on so many different platforms. If you have refinements of these directions for specific platforms, please submit them to mozilla-webtools@mozilla.org Ensure you are running at least MysQL version 3.22.32 or newer. Earlier versions had notable security holes and poorly secured default configuration choices. There is no substitute for understanding the tools on your system! Read The MySQL Privilege System until you can recite it from memory! At the very least, ensure you password the "mysql -u root" account and the "bugs" account, establish grant table rights (consult the Keystone guide in Appendix C: The Bugzilla Database for some easy-to-use details) that do not allow CREATE, DROP, RELOAD, SHUTDOWN, and PROCESS for user "bugs". I wrote up the Keystone advice back when I knew far less about security than I do now : ) Lock down /etc/inetd.conf. Heck, disable inet entirely on this box. It should only listen to port 25 for Sendmail and port 80 for Apache. Do not run Apache as nobody. This will require very lax permissions in your Bugzilla directories. Run it, instead, as a user with a name, set via your httpd.conf file. nobody is a real user on UNIX systems. Having a process run as user id nobody is absolutely no protection against system crackers versus using any other user account. As a general security measure, I recommend you create unique user ID's for each daemon running on your system and, if possible, use "chroot" to jail that process away from the rest of your system. Ensure you have adequate access controls for the $BUGZILLA_HOME/data/ and $BUGZILLA_HOME/shadow/ directories, as well as the $BUGZILLA_HOME/localconfig and $BUGZILLA_HOME/globals.pl files. The localconfig file stores your "bugs" user password, which would be terrible to have in the hands of a criminal, while the "globals.pl" stores some default information regarding your installation which could aid a system cracker. In addition, some files under $BUGZILLA_HOME/data/ store sensitive information, and $BUGZILLA_HOME/shadow/ stores bug information for faster retrieval. If you fail to secure these directories and this file, you will expose bug information to those who may not be allowed to see it. Bugzilla provides default .htaccess files to protect the most common Apache installations. However, you should verify these are adequate according to the site-wide security policy of your web server, and ensure that the .htaccess files are allowed to "override" default permissions set in your Apache configuration files. Covering Apache security is beyond the scope of this Guide; please consult the Apache documentation for details. If you are using a web server that does not support the .htaccess control method, you are at risk! After installing, check to see if you can view the file "localconfig" in your web browser (e.g.: http://bugzilla.mozilla.org/localconfig). If you can read the contents of this file, your web server has not secured your bugzilla directory properly and you must fix this problem before deploying Bugzilla. If, however, it gives you a "Forbidden" error, then it probably respects the .htaccess conventions and you are good to go. On Apache, you can use .htaccess files to protect access to these directories, as outlined in Bug 57161 for the localconfig file, and Bug 65572 for adequate protection in your data/ and shadow/ directories. Note the instructions which follow are Apache-specific. If you use IIS, Netscape, or other non-Apache web servers, please consult your system documentation for how to secure these files from being transmitted to curious users. Place the following text into a file named ".htaccess", readable by your web server, in your $BUGZILLA_HOME/data directory. <Files comments> allow from all </Files> deny from all Place the following text into a file named ".htaccess", readable by your web server, in your $BUGZILLA_HOME/ directory. <Files localconfig> deny from all </Files> allow from all Place the following text into a file named ".htaccess", readable by your web server, in your $BUGZILLA_HOME/shadow directory. deny from all